In 2024, NASA detects a pair of planetoids, each the size of the earth's moon, careening towards our solar system. The good news is that they're set to enter earth's orbit without any collisions (how this happened is the purview of another question). The bad news is that having 2 more moons will cause massive tidal waves, tornadoes, and countless other disasters that will cause human society as we know it to be destroyed.

Needless to say, they freak out and spread the word that a planetary collision is incoming. By some miracle that is beyond the purview of the question, they manage to get the near-total support of everyone in power in the US and UN, successfully convincing anyone who might oppose their measures that it's better to be alive 60 years from now than comfortable right away.

With that out of the way, they immediately set to work on an off-world habitat to which humanity could retreat and wait out the upcoming extinction event. It could be a space station, but it could also be a moon base or mars base. Given 60 years before the planetoids hit and total financial/political support from both the US and UN within reason for that entire time period, Where would they construct this habitat, and how many people could it sustainably support?

Note that the US and UN will not bankrupt themselves over this, but they WILL transition to a war-time economy and give as much financial support as they can afford.

Note also that the disasters are going to affect every inch of the crust, rendering survival on earth anywhere basically impossible for humanity as we know it for at least 30 years, and maybe even multiple human lifetimes. And 'anywhere' includes underground bunkers beneath the surface.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – L.Dutch
    Commented Apr 21, 2022 at 16:17

7 Answers 7


If space has to be the only solution (and as others have pointed out its very tough to get to work), then there is one obvious "best" location.

First, lets consider your resources. It's difficult to predict what launch technologies will look like in the next 50 years. But we do know that reusablity is going to dramatically reduce costs, because it already has.

SPACE WAS EXTREMELY EXPENSIVE. Getting to space has been historically extremely costly. A Shuttle flight cost around $2B in current dollars, and was so costly for two main reasons. One it was a government project built by cost plus contractors where components had to be built in all 50 states (massive solid rockets trucked from Utah to Florida?). More importantly it wasn't really reusable, it required hugely expensive maintenance. The Shuttle engines had to be entirely rebuilt every launch, the solid rocket boosters were destroyed on landing and only parts could be used, and a hugely expensive Hydrogen fuel tank was burned up every flight. Plus the rest of it required expensive maintenance between flights, the Orbiters critical heat tiles had to be carefully inspected and replaced.

The Shuttle was the most expensive launch vehicle ever in cost per pound, around \$40,000 per lb for the payload (ignoring the Orbiter). Thats is the main reason why the ISS cost over \$150B to build. But now a Falcon 9 reusing the first stage can put about 2/3s as much payload in space for \$50M (about \$1,500/lb), and that's at a profit.

REUSABLE ROCKETS WILL MAKE SPACE FLIGHT A LOT MORE LIKE COMMERCIAL AIR TRAVEL Over the next 60 years it's clear that fully reusable space vehicles will dominate space launch. Fully reusable means a rocket that can launch payload to orbit and all of the rockets components will return to earth where they are quickly and inexpensively inspected and refueled to fly again. SpaceX is building a design right now, the Starship, and even if Starship fails more designs are going to be attempted (RocketLabs Neutron is another) until one succeeds. The benefits are simply too massive to ignore. If Starship meets its initial design goals it will cost less than \$30M for 150,000 lbs to orbit, or about \$200/lb.

But how much cheaper can a reusable rocket system get? Essentially when rockets can be reused hundreds of times they will have similar economics to commercial jetliners, where the most expensive cost becomes fuel instead of expending key components every flight (note that similar economics doesn't mean as cheap as). Starship is a Super Heavy launch vehicle, it will be the largest ever made, and its fuel costs are estimated at about \$1M per launch. Musks goal is to launch each Starship one hundred times to spread the costs of construction over many launches. If he achieves that the cost per flight can be as little as \$5M, or \$33 per pound to space.


You might also look into nuclear rockets like the NERVA project successfully test fired in the 1960s, but as I'll explain they aren't likely to offer better economics than a high cadence reusable chemical rocket like Starship for your obvious destinations.

So great, we can lift large payloads and lots of people into space very cheaply now, where do we send them?

THE MOON IS UNINHABITABLE Not the moon. It's an inhospitable desert lacking in resources. Water is only available on the poles and the rest of the moon is over 250 degrees for two weeks at a time, and near absolute zero with no solar power the next two weeks. There are lots of elements in the regolith you could try to melt out, but that's an immense amount of energy. And the regolith is razor sharp, so you must keep it out of crew quarters to avoid breathing it in. The reason the regolith is razor sharp is its never been weathered, it has no atmosphere which means every payload has to use a huge amount of fuel in order to land on the moon.

MARS IS BETTER, BUT STILL NOT THE ANSWER Mars is better. It's much farther than the Moon, about a 3-9 month trip depending upon when you leave and the amount of fuel you spend. But it's atmosphere makes a huge difference. It actually takes less energy to go to the surface of Mars than the Moon because you can aerobrake into its atmosphere without using hardly any fuel. Its temperature range is less than half the moons because of that atmosphere. And it's awash with resources, CO2 for producing fuel and oxygen, ice water for many uses, nickel iron meteorites littering the surface. And the atmosphere has weathered the dust so it's not going to rip apart your lungs or space suit (though it does have slightly poisonous perchlorates that need to be washed off when anyone comes inside).

But Mars isn't the answer either. First, going to Mars or the moon means you increase your fuel requirements as much as ten times from just going to space. The SpaceX plan for exploring Mars with Starships requires as many as ten or more tanker flights to refuel the Starship for the journey. Now your \$30 per pound cost just ballooned back to \$300+. Nuclear Rockets can't help you here, because of their required shape and shielding they can't use aerobraking effectively, so their power advantages are lost to Mars.

And you'd have to send millions of tons of supplies, tools, and equipment to ensure the colonists had everything they needed to last as long as it takes for Earth to return to habitability. They'd have to be self sufficient. How do you make that happen? Maybe they could survive on nuclear power and hydroponic crops grown in underground shelters to avoid the higher surface radiation. But how would they repair things? How would they even make plastics? How would they get all the elements and organic molecules we take for granted on earth?

SPACE OFFERS ONLY ONE CHOICE No you really only have one choice if humanity is forced off planet. Earth orbit. Its by far the cheapest and easiest place to send people. You can put at least ten times as many people in low earth orbit than on Mars or the Moon with the same number of launches. Its protected by the Van Allen belts giving much lower radiation than open space. Its super close meaning your cargo rockets can turn around quickly to fly even multiple times a day, far more often than if they were stuck on week long lunar trips or years long Martian trips. And if the conditions on earth ever improve briefly, say during specific orbital aligments, you'll be able to pop down to collect raw materials.

At \$30 per lb you could launch the million pounds of materials to make another ISS for only \$30M. In reality the ISS is far too small to survive for decades on its own, you'd need to build structures a hundred times larger where a thousand times more people could live. They'd have to be redundant so that any leaks or damage could be isolated and repaired without the entire structure losing atmosphere. And they'd have to be spinning, because humans can't survive in zero gee for years on end without extremely damaging health issues. Essentially you need to build O'Neil Cylinders, large spinning cylinders where people live on the insides. Probably the smallest ONC that could provide for long term survival is at least a billion pounds (ten times the size of a Supercarrier).

And they'd need a ton of redundant power systems. You can't just rely on solar no matter how strong it is, if your panels fade over time you need to be able to replace them or augment them with other power sources (like nuclear).

Devoting 10% of the US Economy to this task for 50 years would give you about \$100 Trillion to invest in it. Using half the funding for payload launch at \$30 per pound, would put 1.5 trillion pounds of space station components (and supplies, tools, equipment, backup supplies/tools/equipment, etc) into orbit. The rest of the money would pay for building the components on earth, flying crews into space to assemble them and finally their inhabitants up to spend their lives in them.

If you assume you need at least 20,000 lbs of structure, hydroponics, materials, tools, equipment, and their backups per person, each billion pound O'Neill cylinder would host 50,000 people, and you could have up to 1,500 of them supporting a total population of 75 million people.

NEAR EARTH ASTERIODS AS RESOURCES You'd also want to investigate near earth asteroids that are within a few months of earth where you can get cheap access to millions of tons of raw resources without having to lift it from the earths surface. They contain water, carbon and metals like iron, gold and platinum. If possible you may want to divert some to orbit next to your habitats so they can be mined easily to augment supplies.

RISKS Lastly you'd still have massive risks. Two new moons would throw the orbits of many near earth objects into widely unpredictable orbits. Massive meteor showers could destroy many of your habitats. Their own orbits won't be stable, they'd have to at least be at the higher parts of the low earth orbits where they won't be in danger of reentering the earths atmosphere for thousands of years. Even then they'll need fuel for orbital adjustments or end up being helpless to avoid any collisions, even with each other.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – L.Dutch
    Commented Apr 21, 2022 at 4:08


The amount of farming land required to support one vegetarian is measured in acres. The cost to launch a kilo of stuff into space is about 10,000 dollarydoos. This is too much to launch into space.

The concept of evacuating to space is silly. When you evacuate you go somewhere safer. Space is the most dangerous place on Earth.

I struggle to think of a level of devastation where you are better off living in a sealed capsule orbiting the planet, compared to a sealed capsule on the planet. This is even ignoring the cost of getting to space in the first place.

Nope. Better to dig under the ground and build an geothermic apocalypse base.

  • 13
    $\begingroup$ +1 for "Space is the most dangerous place on Earth" $\endgroup$ Commented Apr 19, 2022 at 22:10
  • 2
    $\begingroup$ While I agree with your points and conclusion; in the interests of pedanticism, a Falcon 9 costs roughly 1,500/lb for a reusable flight now (50M/17 tons), a reusable Falcon Heavy launch costs just under $1,000/lb (90M/50 tons). Those are prices, actual costs are roughly a third less. If Starship can achieve decent turnaround/refurbishment costs it will cost roughly 30M a launch for 150 tons, or 100/lb (Musk is claiming eventually 5M/flight, but we know how optimistic he is). Doesn't change your conclusions significantly though. $\endgroup$ Commented Apr 19, 2022 at 23:47
  • 4
    $\begingroup$ @SafeFastExpressive Don't listen to Elon Musk. Everyone knows he isn't a real person. $\endgroup$
    – Daron
    Commented Apr 19, 2022 at 23:50
  • 1
    $\begingroup$ The launch cost calculations don't make any sense. If we wanted to construct these kind of habitat stations, you need to consider the Earth to Moon Launch cost of the materials needed to get the Luna industry going to the point where it can launch the bulk materials cheaply. Lunar bunkers still beat space stations in an Earth is lost scenario. But even then, you launch people, high-tech devices and organics from Earth, the rest is produced via in situ resource utilisation. I never get why people ignore that in these kinds of discussions. $\endgroup$ Commented Apr 20, 2022 at 6:32
  • 3
    $\begingroup$ @SafeFastExpressive Your numbers use lb and $. If you convert this to real metric units like kg and dollarydoos like in Darons answer this comes to the exact 10,000 they claim. $\endgroup$
    – quarague
    Commented Apr 20, 2022 at 7:04

From an astronomical point of view there are several scenarios

There are two main possibilities: 1) the objects proceed in an elongated, elliptical orbit and 2) the objects follow a hyperbolical trajectory.

In descending order of disaster level..

Unstable, elongated elliptical orbits in this case, a collision follow, some time in the future. The UN will want to invest a few hundred million of their funds to simulate accurately what is going to happen. If the orbits vary, danger is some day they would collide the planet. In that case, there is no chance of survival on the planet, nor in orbit. Better park your spaceship somewhere at Lagrange L2 and make it a generation ship ! These people will have to survive in space for thousands of years, at least, or settle on the moon, after Earth is changed into a wobbling fireball. No life remains on the planet.

Stable, elongated elliptical orbits along a tilted plane. periodic, huge volcano outbursts, any coastal region will be hit by kilometers of water each tidal cycle. Changes in the atmosphere composition, and a permanent nuclear winter. The dust will eclipse the sun. Underground shelters are no option. Where would these poor underground people go? there will be no safe place left on Earth after several orbits, because the rotation will allow the objects to cause havoc on any location/latitude on Earth. Eventually, the Earth's crust will crumble and disintegrate completely, the heat of the friction causes the crust to partially melt. There is no way for any life on Earth to survive this. In orbit, it could be safer than in the above, unstable scenario, but your spaceship will need a good ion-thruster to compensate for tidal forces while in orbit ! it will need fuel.. and there is little hope of ever returning to the surface. The Earth has become a lava lake with hot islands.

Stable, elongated elliptical orbits, both in Earth's equatorial plane. periodic, huge volcano outbursts, any coastal region will be hit by kilometers of water each tidal cycle. Changes in the atmosphere composition, a Tsunami kilometers in height, and a permanent nuclear winter. The dust will eclipse the sun forever. People can survive in underground shelters in the polar regions (bio-sphere's?), where the floods cannot reach them and tectonic plates are stable (e.g. Antarctic plate). The underground shelters would need a lot of energy to keep the temperature up. On the long term, the Antarctic ice could melt. The issue with any periodic scenario is more and more energy will be dissipated in the crust, causing volcanic lakes along the equator.

One time event, hyperbolical orbits. They come at high velocity, let's hope they'll never return, or be caught by the sun. Along some fault lines and especially in the plane of the trajectories, huge damage to the Earth's crust is caused, there will be abundant volcanism and changes in the atmosphere composition, a nuclear winter. But after this pass by, there will be no more energy dissipated into the crust, after the event. The event would cause unimaginable floods, large part of your underground shelters will be crumbled and cooked in new volcanism, there will be world wide climate change.. most probably an ice age will follow, as a result of CO2, the equatorial region will be burnt to the ground. A mass extinction event, but thousands could survive on the planet surface.. and find a way to rebuild civilization..

  • 1
    $\begingroup$ Two duplicate answers? or am I getting crazy? $\endgroup$
    – Nuclear241
    Commented Apr 20, 2022 at 2:41
  • $\begingroup$ Thx @Nuclear03020704, my last edit seemed to have landed in a separate answer.. corrected. $\endgroup$
    – Goodies
    Commented Apr 20, 2022 at 6:16
  • $\begingroup$ We're dealing with a situation where the planetoids enter near-earth space in what were probably once elongated elliptical orbits in Earth's equatorial plane, but then get caught in earth's gravity (don't ask how, nobody knows why) and stay there. The apocalyptic conditions you described above will exist for a time, but things will eventually stabilize and civilization could rebuild itself, it would just be doing so on a planet with 3 new moons and radically altered tides and terrain as a result. $\endgroup$
    – Brinstar77
    Commented Apr 20, 2022 at 12:04
  • 1
    $\begingroup$ @Brinstar77 that clarification deserves (or deserved) a place in the opening text. It is probably too late for that, because any structural changes in the question will invalidate some answers. $\endgroup$
    – Goodies
    Commented Apr 20, 2022 at 16:51

You Need to go to Mars

There is only one scenario where these new moons could cause a world ending event, and tidal/wind sheer is not it.

To even make this question make any since, lets assume these new moons settled in right at thier Roche Limit which for an Earth/Moon relationship is about 11,470 miles above the Earth. At this range the moons would pull on the surface of the Earth with about 360 times as much force as our current moon. Yes, this would create massive tidal waves ruining coastal cities across the globe, but keep in mind that this is still only about 0.001Gs of force so, not really enough to substantially effect life in-land. The real devastation though of these low orbital moons is from what the Earth will do to them.

The Earth's gravity will slowly pull the moons apart causing a constant and massive meteor shower to pepper the earth. This meteor shower will devastate everything in its path, and kick up enough dust to block out the sun making the Earth uninhabitable for the entire foreseeable future of humanity.

However, these new moons also create such a long-term hazard that you need to make a permanent plan for evacuation. Deep space or the moon will not give you the raw materials you need to maintain a colony for very long without support from Earth. While the moon has raw elements, they are all in undifferentiated regolith which makes it worthless as a place to mine for the things you need. Deep space has the opposite problem. Astreroids often contain large amounts of useful elements, but because each individual asteroid comes from a single type of star death, they tend to lack the diversity of elements you need without have to maintain a massive network of mining ships to go out and mine all the individual asteroids you need (and asteroids are not nearly as closely packed together as Hollywood makes them look). Mar's geological history makes it suitable place to find a wide range of useful ores that can be refined into whatever you need in economically helpful ways. These differences are the whole point of the Space X mission going to Mars instead of the Moon or deep space.


A Falcon Heavy rocket is the most cost efficient rocket we have to date and it can launch about 16.8 tons to mars for a cost of about \$9,077/kg. According to the SpaceX mars mission plan, using light weight inflatable infrastructure where possible, we may be able to colonize Mars at a total mission weight of as little as 3 tons per person. This is also much lighter than what a deep space habitat would be when you consider that you don't need to bring all your water/soil/etc.

So, a Mars evacuation plan would cost about \$27.2 million per person in rocket costs plus probably a few million for all the high-tech infrastructural stuff; so, lets assume a cost of \$30 million per person. That said, if you plan it right you might be able to save even more people if your early colonists can build enough manufacturing capabilities early on, later missions wont have to send the whole 3 tons of stuff per person along. So, you might be able to get it down to closer to 1 ton per person in latter missions when you are only sending people with enough supplies for the trip itself.


The total wartime cost of WWII was about 4 trillion dollars (58 trillion in today's economy). Also consider that the Earth's population was only about 3 billion in WWII, and is about 8 billion now, and is is expected to reach equilibrium at about 9 billion within the next few decades due to shrinking average family's size. So if we consider this level of "war time spending" on a per capita basis over a 60 year period, you are looking at a total budget of about 2.5 quadrillion dollars in today's economy.

This means you can build a self-sufficient Mars colony big enough for somewhere in between 90-270 million people allowing you to evacuate roughly 1-3% of the Earth's total population.

But... there may be another problem here. While the money is in theory here, the natural resources may not be. It takes about 4 units of crude oil to make 1 unit of RP-1/LOX propellant for a falcon heavy engine. Since a falcon heavy has a fuel wight of 92,670 kg and a barrel of crude oil is 136kg, this means each launch requires about 2,725 barrels of crude oil... which is in theory enough to launch 3.6 billion people too Mars with our current fuel reserves; however, this fuel reserve will only last another 46 years at current rates of consumption from all the other things we are doing just living here in the mean time: and this is supposed be a 60 year plan. While you can probably get your first few million people off world at a reasonable price, as world oil reserves deplete, the cost of each mission will get much greater, and by year 45 there may be no oil left to fuel these missions.

You can address this problem in one of 2 ways: #1 assume your program bought up and hoarded the oil you need early on and set it aside for later missions, or #2 assume that the oil crises the world will face in 20-40 years will destroy your economy and effectively cut your evacuation program short. How idealistic you are here will have a significant impact on the size your your mars colony.

enter image description here


With a 60 years heads up, a lot of effort can be devoted to altering the trajectory of the incoming object, some of these methods can be relatively passive like altering the albedo, but even relatively small thrusts can have a large effect.

The cost per kilogram per launch is already dropping (at least for low earth orbit) and will end up be in the 10s or hundreds of dollars per kilogram. If material can be extracted from the moon or asteroids costs continue to drop.

Food supply, radiation, people heath mental and physical are probably limiting in almost all sensations though…


Half a million.

I remember when smoke was rising the second Trade Tower, some of the news announcers started to speculate openly whether it might have been an act of terrorism.

On that note, the pair of planetoids racing straight at Earth might at first arouse only timid consideration of how unlucky we are. But when they start decelerating so they can go into orbit? That would be a clue.

Sending planets at an inhabited planet has a few explanations. Most of them (hungry monsters, borrowing Earth's ocean to refuel) don't allow for a lot of options, and may not have any survivors. So the humans race out with exploratory probes and manned missions.

To get a nonzero result, I'm going to suggest they found the things are meant as an ark of some sort. The aliens have been watching When Worlds Collide, perhaps. So when the first probe does its flyby, it sees a barren moonscape relieved by a fertile river valley, or a hollow sphere full of Nazis, or some other thing suggesting atmosphere and arable land.

The rest is just a matter of building ships - lots of ships. According to StackExchange the cost of an ISS launch is \$55 million an astronaut. Mass production cuts that, but flying to an orbiting planetoid needs more distance, and there's some equipment to lug. So let's say \$50 million per person. The US Federal Reserve estimates \$168 trillion in assets. Assuming 1/3 of these available assets (\$50 trillion) is spent during an exodus, minus 50% for wars with all the other would-be exoduses (we won't count those assets either), we get oh, half a million astronauts. In order to sell this idea to the wealthy and powerful, who control most of those assets, we'll assume that (1) all the astronauts are drawn from their class, (2) the wealthy who don't get to go get clear title to all property, ideas, naming rights, and people left on Earth (who knows, it might survive and they can laugh at the runaways), and (3) a major tech mogul gets to implant microchips to agonize any potential rebels among the slave population.

As no one would seriously criticize a program to give away wealth to the very rich during a national emergency (see also 1, 2), the course is settled. The larger bulk of the evacuees will go to the larger and perhaps safer of the planetoids, with a smaller group as a backup on the other. Maybe 400,000 settle along Musk Canyon on Thiel, and 100,000 in Putin Crater on Bezos.

This count is for the number evacuated, not the number surviving. The fighting among these best of the best of us for toilets, servitor robots, and frozen slave embryos will be colossal.

  • $\begingroup$ Musk canyon is smelly, but not as bad as Poo-tin crater. $\endgroup$
    – Daron
    Commented Apr 21, 2022 at 17:13

Frameshift (but you're not going to like it):

A high tech society (and it has to stay high tech to maintain it's facilities) has a large minimum population requirement due to the fact that people and equipment doesn't come in fractions. You're going to need a lot of specialists and you're going to need backups (what happens when your only widget frobber walks in front of a bus before anyone is trained to take over?) There's no way we are building a self-sustaining offworld habitat of that size in time.

However, you're talking capture in Earth orbit. Something of that size can't aerocapture, the only option is gravity capture by Luna--and note that this must be very gentle even in comparison to normal gravity captures or Luna ends up ejected. The approach velocity has to be very low--but the objects are carrying at a minimum 12.32 km/sec. That has to be shed in other slingshot maneuvers--and note that slingshot maneuvers are very sensitive to the initial approach. This needs multiple slingshots, to hit the last one correctly the previous one needs to be even more precise. (Note that NASA has rockets on craft that will be playing planetary billiards, they can adjust for an imperfect slingshot.)

The velocity change required to disrupt this chain of events is minuscule if done early enough. Instead of a hopeless evacuation the world should be looking to hammer the moons as hard as they can.

(And note the corollary to the high precision required--we simply do not have the ability to determine where they are that precisely in the first place. If they're on such a deadly trajectory we don't know it.)


You must log in to answer this question.

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