I would like to have Earth slowly move away from the sun, taking somewhere between several months and a couple of years to reach the point where human life on the surface is no longer possible. (The story revolves around efforts to get protected in time and the resulting conflicts. So yes, we have some level of spaceflight, but not FTL -- the people who get out will be taking their chances on something like generation ships. Or, as pointed out in the comments, by building shelters on earth to give them longer to come up with a permanent solution.)

My problem is that I don't know what could cause this. I can think of ways to knock the earth out of its orbit quickly that involve killing most or all of the inhabitants, but I need something that leaves many people alive and gives them time to stew about what's next. I want them to be slowly freezing, not slowly cooking -- moving away from, not falling into, the sun. Is this possible? If so how?

It's ok to vaporize a region if that's needed. I'm not fussy about which people survive, only that many do.

  • 1
    $\begingroup$ I'd say that "several months to a couple of years" will be a little bit too ambitious (at our current technological level, the question is a little fuzzy about that). I have to feeling that building one or a few generation ships to get "many survivors out" is more likely going to take a few decades. $\endgroup$
    – Ghanima
    Commented Jan 11, 2015 at 21:45
  • 1
    $\begingroup$ @Ghanima thanks. The idea wasn't to start building generation ships when the problem is discovered; that'd take too long. I'm thinking more in terms of adapting existing shorter-range ships (so a little farther in the future as we'd have ships) quickly as as way of getting some people off the planet. (When I said "something like" generation ships I had in mind this "let's kludge something together out of components already on hand" appraoach.) $\endgroup$ Commented Jan 11, 2015 at 21:48
  • 11
    $\begingroup$ You know, even if Earth was cast into interstellar space and totally froze over, it would still be more habitable any spaceships you could build on short notice. It has gravity, a magnetic field and lots of mass for radiation shielding. Plentiful energy in form of radioactives, hydrocarbons (global warming would not be an issue) and geothermals. Plenty of volatiles such as water and oxygen. Building insulated shelters would be lot easier than building spaceships and result in much larger portion of population surviving. $\endgroup$ Commented Jan 11, 2015 at 22:36
  • 5
    $\begingroup$ So people would build large shelters on the planet to save as many as possible. And then much later then the situation stabilized, start dreaming of finding a new planet with liquid water. $\endgroup$ Commented Jan 11, 2015 at 22:37
  • 1
    $\begingroup$ Well we don't know that much about how stars work, having a freak event cause the radiation output to start falling wouldn't need to be explained. You could even have scientists saying things like "I don't understand how it's happening, it doesn't make sense". $\endgroup$
    – Tim B
    Commented Jan 12, 2015 at 13:20

6 Answers 6


You're going to need a lot of energy. As we'll see in a moment, too much energy to do it quickly.

Assuming we want to reduce incoming solar energy by roughly 20% (should lower mean global temperatures by roughly 15C), we need to increase the earth-sun distance by 10% (incident solar power drops with the square of distance). That means moving the earth 15 million kilometres further from the sun.

The gravitational potential U of a body of mass m at distance r from a mass M is given by $U=\frac{-GMm}{r}$. We add a factor of two to account for the kinetic energy: $U=\frac{-GMm}{2r}$

To get the change in potential, $\Delta U=\frac{-GMm}{2.2r}-\frac{-GMm}{2r}=\frac{.1GMm}{2.2r}$

That will need around $2.5\times 10^{32}$ Joules, which is more power than all the world's reserves of fossil ($4\times10^{24} $J) and nuclear ($2\times10^{23}$J) fuel, or around 50 million years worth of solar energy. It also happens to be around equal to the earth's gravitational binding energy. That's the energy you need to remove every part of the earth from every other part and bring them so far from each other that they won't reform. If you apply this energy in an -even slightly - unstructured way (say, by vaporising a continent), you're not going to be worried about the planet cooling down, you're going to be worried about the fact that the floor is made of lava.

You could have gravitational interaction with one of the gas giants do it, but it will take a lot longer than you want and you'll need to justify why it's only starting now.

Probably your best bet is a rogue planet passing through the system and raising earth's aphelion (you can't switch to a new circular orbit with just one interaction, but an elliptical orbit will do fine. Even at perihelion the earth will be colder than now due to thermal inertia).

  • $\begingroup$ Question out of curiosity: What formula did you use to get that number ($5 \times 10^{32}$)? $\endgroup$
    – HDE 226868
    Commented Jan 11, 2015 at 20:51
  • $\begingroup$ Thanks. For an orbiting body, though, we have to take into account the kinetic energy, too, because bodies with smaller semi-major axes revolve quicker. Because potential energy is negative while kinetic energy is positive, the formula becomes $$U=-G\frac{Mm}{2a}$$ where $a$ is the semi-major axis. $\endgroup$
    – HDE 226868
    Commented Jan 11, 2015 at 21:16
  • 1
    $\begingroup$ @frodoskywalker Is there any reason you couldn't just have a rogue planet come in and set Earth on a hyperbolic trajectory out of the solar system, or raise Earth's Aphelion so much so that it's effectively uninhabitable? Time scales could be as long or as short as needed, depending on the scale of the interaction and how much warning the people on Earth had. $\endgroup$ Commented Jan 11, 2015 at 22:09
  • $\begingroup$ @cartographer nope, and that's a lot more realistic than my scenario since it doesn't require a quick capture of the rogue planet. $\endgroup$ Commented Jan 11, 2015 at 22:12
  • $\begingroup$ A Pail of Air by Fritz Leiber (1951) posits that Earth was captured by a passing "dark star". And I've just read "Star with potential to disrupt solar system identified" < wired.co.uk/news/archive/2015-01/05/… >. $\endgroup$
    – mirimir
    Commented Jan 12, 2015 at 0:02

Roaming around interstellar space are rogue planets and rogue stars. These are planets and stars that have been thrown out of the star systems where they were formed. I have seen an estimate that says that as much as 50% of the planets formed within a star system are ejected by the time the system matures.

If one of these planets or stars swept through our solar system, a number of things could happen.

  1. The path of our star (the Sun) could be perturbed. There are many ways in which the Sun's path could be perturbed. Worst case, it could simply be sent careening away from the planets and the planets, no longer having something to orbit would begin travelling in a straight line along their current path. The lights would go out on Earth pretty quickly in this case: days/weeks. A more "mild" case of the Sun's path being perturbed could shift its trajectory in such a way that the Sun slowly begins moving through the orbital disk of the planets. In this case, depending on the speed of the Sun's progression, the planets might slam into the Sun or they might pass very near the Sun and be thrown into highly elliptical orbits.
  2. If the rogue element passes through some part of the orbital disk of the planets around the Sun, it could perturb the orbit any planets it passes near. For example, if a very large rogue planet passes behind Mars in its orbit, it could slow Mars down such that Mars' orbit could become elliptical and at its perihelion its orbit could be inside the Earth's orbit. If this were to occur, it's possible that we could use orbital projections to determine that in 3 or 30 or 300 years, the Earth and Mars would collide.
  3. If the rogue body passes near Earth, the Earth could be sped up or slowed down in its orbit, in either case causing an elliptical orbit. If the Earth is sped up, it would go through periods of substantial cooling during aphelion when the Earth was farther away from the Sun. If the Earth was slowed down, it would go through periods of substantial heating when at perihelion the Earth was much closer to the Sun.
  4. The rogue body could pass inside the Oort cloud way out at the extreme edge of our solar system. The Oort cloud is thought to contain comets and planetesimals left over from the formation of the solar system. Most of the comets, asteroids, and other leftovers from the early solar system have either collided with the Sun, the planets, or have been ejected from our solar system. Thus our current solar system is "relatively" uneventful in terms of heavy collisions. Something passing inside of the Oort cloud could dump a very large amount of that debris into the inner solar system which depending on the amount material dumped could overwhelm our ability to even detect if something were on a collision course with Earth. It would take quite a long time for this material to approach the Earth and we would see a lot of it coming for at least a decade, probably 3-5 decades. And as the material passed by any outer planets on the way in, those planets would fling the material off in various new directions. And some of it would slam into the outer planets and create impacts as we saw with Shoemaker-Levy and Jupiter. So there would be a lot of new stuff for the Earth to run into for a few 100 or 1000 years. So that would suck...

It's possible that we could detect the rogue element approaching our system and predict its effects relatively far in advance. For literary purposes, it could be decades, or we could detect it in the last few years of its approach (in the case of an Oort cloud intersection) leaving us little time to react.


When you want the orbit of earth around the sun to get higher, earth isn't falling out of orbit, it is raising out of orbit. To get an object in orbit to a higher orbit, you need to add rotation energy to it. That energy needs to come from somewhere.

A massive object passing through the solar system and disturbing orbits (as Kent described) would be an option, but it would not be a continuous development but a rather sudden event. Also, it likely would put earth on an excentric orbit and not a round one like it currently has.

Another option would be to leave Earth alone and instead reduce the mass of the sun. When the sun would gradually lose mass, its gravity would also be reduced and all planets would go onto higher orbits. However, whatever causes the sun to lose mass would likely also affect its energy output in one way or another. So the main problem for life on Earth would likely not be Earths changing orbit but rather the sun becomming hotter or colder than usual.

This leaves the question: What could cause the sun to suddenly start losing a significant amount of mass?

One possible path how the sun could lose mass would be through a sudden increase of coronal mass ejections (preferably in polar direction so the resulting solar winds don't destroy all life on earth). Truth be told, we don't actually know much about what is going on inside a star. Most of what we believe to know is based on unconfirmed hypotheses. So you can get away with quite a lot of plausible explanations why this is suddenly happening.

Another option would be teleportation (when you want to allow it - it's a very soft sci-fi trope). Something or someone teleports large amounts of mass from inside the sun to somewhere else.

  • 1
    $\begingroup$ +1 for reducing the mass of the sun, which is more easily explained than the earth changing significantly. The sun could get both slightly colder and have a reduced gravity, so you get double the effect for a fairly small change. $\endgroup$
    – Jon Story
    Commented Jan 12, 2015 at 13:44
  • $\begingroup$ An alternative to changing the mass of the sun is to have some event change the gravitational 'constant', probably transitioning slowly, with some acceleration. It could be detected as a result of experiments in detecting gravitational waves, giving some advance notice that there is an accelerating change that will soon result in some major issues for Earth. $\endgroup$
    – Dan Bryant
    Commented Jan 12, 2015 at 16:53
  • $\begingroup$ @DanBryant Changing the gravitational constant would cause major issues in the whole universe. It will, for example, destabilize the stars, because every star has a delicate balance between fusion pressure from the inside and gravity pressure from the outside. $\endgroup$
    – Philipp
    Commented Jan 12, 2015 at 17:46

Do you specifically have to move the planet?

If not, what's wrong with a good old giant-asteroid-mass-extinction-event like the one that did for the dinosaurs?

A big comet smashes into the world throws up a load of dust and crud into the atmosphere which prevents the sun's light coming through, temperatures drop, plants stop growing etc.

You could even have an giant-asteroid shower of sorts which lasts several years, causing chaos and slowly making things worse as these giant asteroids smash into the planet every few months, destroying a new part (making staying risky even if the dust/temperature drop doesn't kill everyone) and throwing yet more dust into the sky.

The other advantage being that there isn't much science to make up or explain.

  • 3
    $\begingroup$ The problem with an impact is that after a few years things start to return to normal. $\endgroup$
    – Tim B
    Commented Jan 12, 2015 at 14:01
  • $\begingroup$ It provides an answer to the spirit of the problem provided, if not the precise wording of the question..... $\endgroup$
    – Jon Story
    Commented Jan 12, 2015 at 15:54
  • $\begingroup$ How long do mass-extinction events take to cause the mass extinctions? If a giant asteroid slams into earth today, will anybody still be here next month? $\endgroup$ Commented Jan 12, 2015 at 22:02
  • $\begingroup$ That depends how big it is, how much dust it throws up etc. a massive one would be instant, but it can be small enough that the shockwaves doesn't kill everyone, but still throw enough dust into the atmosphere to block the sun. So it would take a while for the plants to die from lack of sunlight, followed by the animals...then we've got a period of killing each other for the last of the tinned food before we die too. $\endgroup$
    – Jon Story
    Commented Jan 12, 2015 at 22:42

I haven't done any calculations on this, but I believe that the Earth and the Moon should be considered one body of mass when calculating its solar orbit.

So if the Moon was perhaps vaporized, e.g. with a huge nuclear device, perhaps the loss of mass could be sufficient for the Earth to escape the Sun's gravity. But I have no idea how long would it take? But given you remove 1/7th of the total mass of the system, I reckon it will be relatively quick.

Or maybe, if the idea is sound, you could first get the Moon to escape Earths orbit in some way. Whatever amount of energy would be needed to get the Earth to escape the Sun, only 1/6th would be needed to get the Moon to escape the Earth. And if such an event was the collision with a very large stellar body, the collision itself wouldn't itself kill off humanity.

This will also give you more freedom in deciding how long a warning time Humanity should have, as you can decide at which rate the Moon should escape from the Earth.


I realized that this answer must be wrong. If we reduce the mass of the orbiting system, it would of course reduce the centripetal force, but due to the reduced mass of the system, the acceleration toward the Sun would be unchanged (just like two objects with different mass fall with the same speed in vacuum)

p.s. Completely unrelated to the actual answer (I don't have enough rep to add comments), Arthur C. Clarke's book The Songs of Distant Earth is similar to your plot description, i.e. humanity trying to create a future for itself as the Sun is about to go supernova using sub-FTL technology

  • 1
    $\begingroup$ If you could do some research and try to answer the questions you raise in this, this would be a top class answer. Welcome to Worldbuilding! $\endgroup$
    – ArtOfCode
    Commented Jan 12, 2015 at 12:53
  • $\begingroup$ There is an issue with "vaporizing" the moon and that is the principle of mass conservation. Unless the particles of the destroyed moon are accelerated beyond escape velocity they will keep orbiting earth (and form some nice rings), but the mass of the system earth-moon will stay the same. The point being it's not sufficient to destroy the moon but to get rid of it. $\endgroup$
    – Ghanima
    Commented Jan 12, 2015 at 13:28

If you found a logical reason to reduce the sun's mass by a significant amount, it would lose its gravitational strength and earth would likely slowly orbit outward until escaping or at least obtaining an uninhabitable orbit.

  • $\begingroup$ Philipp’s answer already includes “reduce the mass of the sun”. $\endgroup$
    – JDługosz
    Commented Jun 1, 2017 at 4:38

You must log in to answer this question.

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