I'm attempting to figure out the orbital physics of a system that has a habitable body that has intense periods of waxing and waning sunlight. I've come to the assumption that the best body for this to happen on would be the moon of a gas giant as I can use the gas giant to periodically shield the planet from star.

This is not something I'm an expert in at all so hoping for help in figuring out what it would take for this scenario to be possible. If it's possible at all.

So here's a lot of information on what type of system I'm thinking about, more details will be added as they come up.

Goals for the moon :

  • Hemisphere A is habitable all the time, even if the environment gets harsh at times, the colonists have means to deal with temperatures up to 60-70 degrees Celsius. Majority of the primary colony and outliers are within exposed cave systems.
  • Atmosphere is breathable by humans
  • Gravity within 10% of earths, size doesn't matter much to me but should support cave systems
  • Developed basic life - mosses, fungus, primitive leafing things, insects, and amphibians. So the system in place has to have been stable enough to allow for life to develop.
  • It's not the only moon around the planet, but only one that has developed life and can sustain humans without creating habitats.

Creating "Sunfall"

  • The moon should go through periods of time say 100-200 years where humans and other life can spread to 60-100% of the moon. Temperatures should reach points where no type of environmental suits are required.
  • The moon needs to go into Sunfall periods of 40-100 years where greater than 50% of the moon will be cooked. Sustained temperatures for long periods of time that will cause plant life to burn/die and would be lethal to humans exposed for more than 5-10min, either due to heat or radiation.
  • Sunfall isn't required to be constant, but last long enough and happen enough that it's not feasible for life to come back until Sunfalls stop, or their frequency drops to a certain point.

Flexible with :

  • It doesn't 100% have to be a moon, I'm just not sure how you'd do this with a planet...would be incredibly awkward combination of things.
  • Environment can be extreme, but humans should be able to survive in the safe region without much in the way of equipment.
  • Colonists are used to weird days and living in cavern systems, they have clocks based on far-flung home planet so rotational and orbital periods don't matter as long as they work.
  • The type of star, or size and composition of the parent planet
  • Cooler periods can be longer, but I'd like to not have Sunfall last more than 1-2 generations of humans.

What would a system look like to get phenomenon like "Sunfall" on a celestial body? If this doesn't seem possible that's also something I'm willing to accept and consider alternate possibilities giving similar impacts on the environment.

  • $\begingroup$ I think your best bet is to put the whole moon/planet on an elliptical orbit around the sun. Any stable orbit around the gas giant would spend very little time comparatively in shadow, not the decades you need. $\endgroup$
    – BMF
    Oct 29, 2019 at 16:13
  • $\begingroup$ I think my biggest issue here is how to not go between Ice Cube and Fireball. Yet have only a very small region that can sustain a constant population, while the rest changes. Period of "Sunfall" being an important factor to the whole thing. An elliptical orbit on its own probably wouldn't do that. $\endgroup$
    – Nymn
    Oct 29, 2019 at 16:25
  • $\begingroup$ Having investigated the concept of gas giant planets causing eclipses on habitable moons, I submit that it would not be suitable for your purposes as described in your question. I asked a similar question here: worldbuilding.stackexchange.com/questions/148473/… . In summary, the longest eclipses between gas giants and moons are on the order of hours, maybe a day at most, if the moons rotation is anywhere near an Earth day cycle. Certainly long enough to cause some severe weather for day from the cooling airflows, but nothing on the scale you need $\endgroup$
    – Harthag
    Oct 30, 2019 at 13:00
  • 1
    $\begingroup$ ...cont. Your description of "waxing and waning" best matches an eliptical orbit, in my opinion, despite your comments on an answer related to that scenario. I'm not familiar enough with variable stars to know, for sure, if that would be a better match, but I've investigated stars enough to know that, in general, variable stars are not friendly to life developing in any form (though I'm not sure a highly elliptical orbit is any better). This might be useful on that front: worldbuilding.stackexchange.com/questions/9857/… $\endgroup$
    – Harthag
    Oct 30, 2019 at 13:02
  • 1
    $\begingroup$ I'll try to explain, without using an earthlike day and instead use the overall orbit of the moon around the planet for reference instead. The moon in my question took just over 1000 hours to orbit the planet, and only 4-6 of those hours were eclipse. That's less than 1% of the time. While you can adjust the parameters of the orbit, and make the planet bigger, so it blocks more light, you'll get maybe a half percent more, at most, maybe 10 or 15 hours, out of a thousand, regardless of how long the day/night cycle is. It'll be nothing more severe than a freak snow in June $\endgroup$
    – Harthag
    Oct 30, 2019 at 13:17

2 Answers 2


Off the top of my head, I feel like the easiest way to handle this is with a binary star system.

Put your habitable planet in close orbit around a dwarf star. Planets close enough to a red dwarf to be habitable at all would almost certainly be tidally locked, which solves for the 'only one hemisphere is normally habitable' criteria in your question. Then have your dwarf star itself be in a highly elliptical orbit around a larger, brighter companion.

I did some calculations and if we assume a fairly standard dwarf star mass (.15 solar masses) with an earth-sized planet orbiting it at a distance that would allow liquid water (.01 AU), the orbital period of the planet around the dwarf would be pretty close to 24 hours.

This is handy because it also means that your planet is going to be rotating quickly enough to maintain a proper electromagnetic field. During most of the orbital period of the smaller star around the larger, your planet is only going to be getting energy from the dwarf, and only on the one hemisphere due to the tidal locking. You're also going to be getting a lot of tidal flexing on your planet which means plenty of tectonic activity, which will help keep things warmer than they might be otherwise.

So. During most of the dwarf's orbit around the companion, your 'day' side will be temperate at the 'east' pole (the point directly facing the dwarf), fading towards an arctic climate as you get towards the edge of the illuminated surface of the eastern hemisphere. Your 'western' or 'night side' hemisphere will be mostly frozen. Think Antarctic winter.

As the dwarf approaches its larger companion though, things will start warming up. You'll have a gradual period where the entire planet starts getting more and more energy from the larger companion and, because of the orbital period, you'll start to have something like a normal 24 day/night cycle increasing in intensity as the dwarf gets closer and closer to the brighter star. This will create your period of global habitability.

At the point of closest approach, your planet will be getting SO much more energy from the bright companion that everybody has to return to the caves until you pass perigee and the dwarf starts its trek back outwards. This is your 'Sunfall' period where the combined intensity of both stars in close proximity creates too much heat for comfort on the surface.

Then you have a second 'season' of global habitability as things gradually cool off, but the planet is still getting daylight from the large companion throughout its orbit.

After a while the dwarf passes far enough from the companion that the energy becomes insignificant again and you have the long cold winter until sunfall comes again.

The more elliptical the orbit is, the longer you'll spend in the 'winter' phase where your planet is only relying on the dwarf, and the shorter your habitable and Sunfall phases will be. Generating the kind of timeframe specified would require an orbital period of the dwarf around the larger star of around 400 years, which requires a semi-major axis of about 60 AU. For comparison Pluto's semi-major axis around the sun is 40 AU. The math of exactly how long the Sunfall period would be in this case is a bit beyond my mathematical capacity so I'm hoping someone else will fill in the gaps on that, but with an elliptical enough orbit the 20-40 years you specified should be easily possible.

  • $\begingroup$ This answer seems perfect, I was going to pen my own also using an elliptical orbit but realized you'd need a supplementary heat source then looked at your answer and it was doing exactly that. $\endgroup$
    – Tim B
    Oct 29, 2019 at 17:15
  • 1
    $\begingroup$ You could potentially do this using a gas giant rather than a 2nd star too - if you get enough heating from tidal flux and the gas giant, etc to keep you from freezing through the winter phase. The binary star solution is less complicated to set up with the right balance though. $\endgroup$
    – Tim B
    Oct 29, 2019 at 17:16
  • $\begingroup$ @TimB it was already on my mind because of the OTHER question from yesterday with a similar kind of problem and, yes, you could do it with a gas giant, depending on how cold you wanted things to be during the 'winter' phases. If you did it that way the planet would have a climate like that of Ganymede or Europa which isn't EXACTLY balmy. $\endgroup$ Oct 29, 2019 at 17:16
  • 1
    $\begingroup$ See Brian Aldiss's Helliconia series for a binary star system producing almost exactly the setup you're looking for, although with somewhat reversed seasons: life is good during Great Summer (close approach to the primary star) and very harsh and cold during the Great Winter. $\endgroup$ Oct 30, 2019 at 8:04
  • $\begingroup$ I'm not a huge astronomy person, but by the orbital period you're meaning it would go around the dwarf every 24 hours, or the backside would basically day/night in that duration? That would work for my purposes. This would basically lead to a planet that basically has an eternal day on one side and a day night cycle on the other one? Due to it periodically not facing the second star if I'm understanding it correctly. Trying to picture how light would be working on the planet with two stars going. $\endgroup$
    – Nymn
    Oct 30, 2019 at 12:11

Variable star.

Your moon is just a humble moon, doing moonly things. Your star, however, is moody. It has dim and bright cycles. It is a variable star.

Periodic measurements of variable stars using AASO data

Variable stars are stars that vary in brightness over time. This can be caused by a variety of effects.... There are also intrinsic variable stars that are caused by changes within the star itself. An example of this is a star that is periodically collapsing. As it collapses the pressure in temperature inside the star increase, causing more light to be emitted. The increasing pressure then causes the star to expand outwards, restarting a cycle.

There have not been accurate astronomical observations for long enough to determine if there really are any stars that cycle with the long period you want. But stars can cycle and making the Sunfall happen because of the star itself would be a fine and plausible thing.

If it is the star that is up to these games, I could imagine that the gas giant would also change in appearance from hot to cold cycle. Its atmosphere would expand as it heated up, for sure. More energy in the atmosphere might also cause electrical effects. I do not think those things would affect a moon in orbit, but you could definitely see them happening from the vantage point of the moon.

  • $\begingroup$ That's an interesting take on it. Makes me wonder if there is a possibility of electrical effects reaching the moons and what effect it would have on them. Will have to research that a bit! $\endgroup$
    – Nymn
    Oct 30, 2019 at 12:07

You must log in to answer this question.

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