Hello I imagine this is a pretty straightforward question, perhaps already covered in another post, but I wanted to get extra perspectives and options for achieving this type of planet or moon. So yeah the topic in question is how a planet or moon could have a variable day/night cycle throughout its solar year.

Example: I'm looking for a planet or moon where one period of the year it is day for 48 hours and night for another 48 afterwards. Then in another period it could be 23 earth days of daytime then 23 earth days of nighttime.

I am looking for something that is very reliable and consistent and it has to still allow for the planet or moon to be relatively habitable at least in certain areas or during certain times in the yearly cycle. The idea would be similar to how some parts of the year up far north have long periods of almost total night or total day. However this would be found in other latitudes and the length of day versus length of night would stay equal. The important part is that this would be very consistent so any civilizations could actually map out when the Ever-Day (very long day) or Trailing-Night (very long night) would occur. Eclipsing effects to achieve this effect are ok, but if there is a way to do this without having the planetoid's star obstructed by another body that would be preferred. Elliptical orbits I imagine are the way to go with tidal locking playing some kind of role, but I don't really know. Any avenues no matter how out there are encouraged, but this is meant to be a scientifically plausible world not reliant on hand-waving magic stuff to explain everything.

I look forward to any of your creative ideas and thank each of you in advance for your help.

Edit: Never mind I underestimated this concept, ignore my initial statement this is not a straightforward question my apologize. However this makes it even more interesting!

  • 1
    $\begingroup$ This looks tough. The only thing I could think of is to play with orbit eccentricity, but I don't think that would give anywhere close to the variation you're looking for. Looking forward to any solutions people find! $\endgroup$ – Qami Oct 28 '19 at 18:42
  • 2
    $\begingroup$ There is some bizarre orbital solution to this involving a binary or possibly trinary star and several other dancing celestial bodies. More than my computer can handle, I’m out! (By the way Venus has a variable length day.) $\endgroup$ – Vogon Poet Oct 28 '19 at 18:45
  • $\begingroup$ I do not think it is possible to have a survivable planet that has the characteristics you want without using a debris cloud. I don't think that it is possible to have a planet that is tidally locked some of the time. $\endgroup$ – Garret Gang Oct 28 '19 at 18:48
  • $\begingroup$ One other effect that can change your planet’s rotation is by changing the spin of the core. If the core accelerates, the surface decelerates. If the core slows down, the surface accelerates. So a planet-sized magneto in résonance might do this. $\endgroup$ – Vogon Poet Oct 28 '19 at 18:54
  • 2
    $\begingroup$ Not really a full answer, but you could get most of this by just having a slowly rotating planet with high axial tilt. E.g: if you had a "day" length of 23 days with an axial tilt of 75 degrees, you'd have both 23-day "days" for most of the year with a long "summer" day whose length would depend on latitude and a similar "winter" night, both of which would be present on most of the planet. Earth-like seasonal changes in day length would give you 45-day "days" just before summer, with 1-day "nights". $\endgroup$ – ckersch Oct 28 '19 at 21:11

if there is a way to do this without having the planet's star obstructed by another body that would be preferred.

I don't really see how, tbh. The problem is that even with a heavily elliptical orbit, your planet's rotation is always the same. An heavily elliptical orbit would certainly give you a wide variation in seasons depending on distance, but your day/night cycle would always be fixed with the planet's rotational period and nothing you do with orbits can change that.

The only thing you can do to make the day/night cycle vary like this is decouple it from the planet's rotation, which means you need a more complex relationship than just "Planet orbiting a star". "Planet orbiting a gas giant" could work quite nicely.

The scenario that seems most plausible to me would be to have a habitable planet orbiting a gas giant that has a dramatic axial tilt relative to it's orbital path around the star. Imagine taking Jupiter (or Saturn, if you prefer) and tilting its axis of rotation (and that of it's moons) 90 degrees. You would have two points in the gas giant's orbit where the axis of rotation would be pointed directly at the star, and thus one hemisphere or the other of the satellite planet would be receiving sunlight continuously (call these the solstices). At two more points of the gas giant's orbit the axis of rotation would be parallel with the gas giant's orbit, and you would have day/night cycle that's directly tied to the rotational period of the satellite planet. (call these the Equinoxes) If the moon is tidally locked then whichever hemisphere is pointed at the gas giant is going to have regular eclipses as it will be passing 'behind' the gas giant relative to the star.

You'd slowly cycle from each condition to the next as the gas giant orbited its star.

For reference, if you assume Jupiter's values, one cycle would take a bit less than 12 years. Your satellite moon would have a couple years of arctic winter in the northern hemisphere, followed by four years of transition to a summer of continuous daylight, and then the process reverses itself.

Depending on proximity to the Gas Giant, your satellite moon could have an orbital period (and thus a day/night cycle) of anywhere between 40 and 400 hours if you want to keep it within the same kind of orbital distance as the Gallilean moons.

  • $\begingroup$ With a highly elliptical orbit, isn't it possible that the closest approach to the star could make for an extended daytime on one side of the planet, if the angular velocity of the planet's rotation is close to the angular velocity of its orbit at perihelion? Effectively, the planet would behave like a tidally locked planet in that part of its orbit. Likely that even if this was the case, the orbital eccentricity would make the planet unbearably hot at perihelion, or frozen at aphelion (or both), but it should satisfy the conditions of the question. $\endgroup$ – asgallant Oct 29 '19 at 16:53
  • $\begingroup$ @asgallant no, because remember, the orbital plane of the moon is 90 degrees offset from the orbital plane of the gas giant, so you can't have that kind of effect. $\endgroup$ – Morris The Cat Oct 29 '19 at 17:00
  • $\begingroup$ I'm not sure what you mean by that, but in any event, the question allows for a planet that has the desired effect, so excluding a moon doesn't invalidate the premise. $\endgroup$ – asgallant Oct 29 '19 at 17:12
  • $\begingroup$ I very much like this approach to the question, I think it makes for a very unique world. I think I will be making use of this planetary system in my story. I'll definitely have to research what a habitable moon of a gas giant would be like. Thank you very much @MorrisTheCat ! $\endgroup$ – Aceharmsway Nov 11 '19 at 18:05

This is sort-of what happens when you have a planet orbiting a single star in a binary system.

In the Illustration below, let's assume the green planet has a 96 hr rotational period.

In position #2, the stars line up and you have a 96hr day.

In position #4, you have two Noon times and 2 twilights giving you sort of two 48 hr days. If your planet has an atmosphere that creates particularly strong rayleigh scattering, if your species can only see in the more reflected wavelengths, or if they just have poor low light vision, these twilight periods could appear like night time.

This may not be exactly what you want though. In positions #1 and #3, you will have an extended day cycle and a shortened night cycle; so, the duration of the day will not scale uniformly, rather it will be an anomaly of a certain time of the year that you get 2 apparent days per rotation.

enter image description here

Another option that fits better in some ways but worse in others would be a tidally locked moon on an irregular, not fully true orbit. This can give you the day length variations you are talking about but these orbits are unstable. No planet or moon would ever form on such a trajectory, but if you have a dwarf planet or large asteroid that gets swept up in the gravity of a larger passing world, its initial orbit will be erratic, and it will take a long time for it to settle into a stable orbit. As the orbit begins to settle, it will begin to develop an increasing level of predictability, this means that you may develop a pattern of orbits where some take much longer than others as it in some cases passes in close and in others swings out wide.

In geological or evolutionary time this would be a temporary condition, but in the lifespan of a civilization, this could become all they've ever known.

One caveat with this kind of situation is that these variations will have nothing to do with time of year. You could have several "day seasons" in a year, or several years in a "day season" just depending. Another is that 48-96hr is a pretty fast lunar orbit. Your best bet to achieve this kind of orbital speed is to be in a relatively close orbit to a gas giant. At the required distance, your planet would experience tidal forces over 10,000 times of that which the moon has on Earth. Being tidally locked will hopefully prevent the worst of the damage that this would otherwise do, but since you are on such an irratice orbit, your tidal lock is not going to be perfect; so, the wobble of your moon relative to the plant is likely to cause some serious tides.

enter image description here

  • $\begingroup$ Your first approach has a simple yet effective method to producing the scenario, however seems to only really allow for extend day periods not as much extended nights. The second method is very cool and something I would not have thought about but the unstable nature may make it not very conducive for intelligent life to have already evolved there. However it is good nonetheless thank you for the help @Nosajimiki ! $\endgroup$ – Aceharmsway Nov 11 '19 at 18:11

The largest problem with changing the day-night cycle duration is that rotating planets represent a massive amount of angular momentum, and making changes to that typically involves 'unpleasant things' having to be done to the surface.

A highly elliptic orbit can induce heavy seasonal influence on a global scale, as opposed to earth's seasonal influence on a hemispherical scale, but you're not really going to easily impact your day-night cycle with one. Instead you'll have extremely hot [potentially dangerously so] days on the short end of your orbit, while having extremely dim and cold days on the long end of your orbit, but the time from sunrise to sunrise remains the same general average.

The easiest way to change the solar day duration for the surface of a planet is to uncouple the surface from the core, and introduce additional mechanics - In other words, don't make it a planet, make it a planet scale machine.

Your surface isn't just the crusty cool rock layer floating over a molten core slowly marching its way towards the eventual heat death of the universe, it is a fly wheel of a giant machine...

Exactly what the machine does, or even who built it, isn't overly important, at least to any civilization who might arise on the outer surface of this "fly-shell", and how much they understand of it isn't overly important to the machine itself. All we really need to care about is that it transfers momentum back and forth between its shell and core while following a consistent process on a regular schedule.

Those on the surface might observe some odd effects beyond the change of day-night cycle duration, and possibly some electromagnetic emissions, but our machine can merrily chug along for eons doing its thing while scientists on the surface debate the nature and structure of what's hidden inside their 'planet'.

[Maybe part of our world building is that some overly clever person or group of people suspects that maybe there is more to the planet than molten rock, or maybe everyone just shrugs and goes on about their lives with fancier clocks and the mindset of "Well that's just the way things are..." with mild suspicions of it being kind of weird.]

Whether the machine-planet is an intergalactic dooms day device or merely a failing grade science fair bauble of a hyper advanced civilization is far less important to us than the fact that the outer shell exchanges angular momentum back and forth on a regular basis...

Other 'slightly' more plausible options may involve getting weird with the material and structure of the planet's core [And thereby having to put up with some community members raising a stink about the planet no longer counting as a planet according to IAU...] such that its rotational properties are not entirely stable.

[While probability of many suggestions in various answers may strongly approach zero in our real universe, we don't have to settle for such boring realities in Fiction...]


Several factors can change your very complicated planet’s day cycle through the year. All of these require your planet to have a shifting mass area, and for your particular case, a very large mass area. Venus has a variable length day by 7 minutes, and part of the explanation is that mountains stop the flow of clouds (mass stops rotating). That only accounts for 2 minutes, the rest is a mystery.

Dynamic braking of the magnetosphere:

Your planet’s core is molten iron with a very unique cycle which repeats every year. This planet is a huge magnet like mercury which generates a magnetosphere from the solar winds.

Mercury's magnétopause

What makes your system even more complex is the sister planet orbiting it (or a very ferrous moon), drawing some of the magnetic flux away for certain parts of the year, which changes the shape of the magnetosphere. The magnetosphere and planet core form a motor which accelerates and decelerates the planet through the year.

Also, the planet’s liquid core is rotating at a different rate than the crust. Both core and crust are heavily ferromagnetic, and they contribute to the motor action.

The timing of all this flowing material, binary orbits, and solar winds is beyond miraculous, but the existence of it is not beyond what is physically possible.

Calculations for a system like this are beyond what we understand today. The answer: Magnets.

I defined “fairly habitable” as simply meaning it has a somewhat constant orbit. And much of the planet’s surface would normally be ripping itself apart, so you’ll need a new question to see if it’s possible to control the tectonic activity on such a planet, with some sort of amazingly smooth and uniform liquid lower mantle. The gravity of a sister planet or moon makes this even more problematic.

  • 1
    $\begingroup$ The amount of energy necessary for even a milissecond in change of day length on an Earth-like planet is so huge that having it happen in one year could melt the crust. This is possible but does not bode well for life, also I think you'd need the star to be a magnetar. $\endgroup$ – Renan Oct 28 '19 at 19:40
  • $\begingroup$ Only if we assume an earth-like life and year. I completely agree the energy used in accelerating and decelerating the planet’s crust would be immense, but the OP didn’t constrain the year length or life conditions. Without a highly elliptical orbit with a sister celestial body, dynamic breaking and motor action are the only forces which can change the rotation rate. Solar winds need to be massive - yeah, a magnetar would do it. $\endgroup$ – Vogon Poet Oct 28 '19 at 19:45

Your planet/moon is small, so it has not pulled itself into a sphere. It is lumpy. As such, under the gravitational influence of other celestial bodies, its rotation is chaotic. While the angular momentum of the world is constant, its axis of rotation changes constantly, as does its moment of inertia and thus length of day.

There are several moon in the solar system that work like this. Hyperion's rotation is so chaotic that NASA could not reliably schedule passes of the Cassini probe to scan new areas of the surface. Pluto's moon Nyx can have days where the sun rises in the east and sets in the north.

It probably isn't possible to predict day lengths for all time, but it would be believable to be able to forecast out a few years.


Obliquity Oscillation

Not all days on Earth are created equal. The days are longer in the summer and shorter in the winter. The reason for this is simple - Earth spins on an axis which is an average of 23.4 degrees. The measurement for this axis is the difference between the rotational plane of the Earth around the Sun, and the Earth's rotation. This is called obliquity.

And the obliquity oscillates. Not a lot on Earth, and very gradually, a very small amount to the point that you barely notice it. This is because it's been stabilize by, among other things, the moon. But it could be much greater, and happen faster. If it does, then it's possible for locations on said planet to be pulled above or below the tilt and thus act as the arctic circle might - have days of sunlight, then days of darkness, and when the obliquity oscillation doesn't pull it in and effect it, normal days.

Of course, there are problems. This would only affect certain areas of the planet, it'd be hard to get to the equator, a destabilized obliquity would play havoc with the weather, and it would take a nightmare's worth of calculations to figure out how exactly it might work. But it's a thought.

  • $\begingroup$ I was thinking something like this, but I would have used the term "wobble". And you put it better than I would have anyway :) $\endgroup$ – Paul TIKI -Monica come Home Oct 29 '19 at 20:49

Hyperion, Saturn's moon, has neither a predictable length of day nor direction of rotation

Hyperion, one of Saturn's moons, fills some of your requirements. But it has a number of weird factors. See https://en.wikipedia.org/wiki/Hyperion_(moon) and https://www.space.com/20770-hyperion-moon.html .

Elongated shape, so that the ends are much farther from the center than the lowlands. Low mass, such that a catapult could fling you off, especially if you started at one of the ends. Very odd composition, probably water ice and organics, with a very high percentage of empty bubble spaces, like Styrofoam. Elongated orbit around Saturn, synced with the larger moon Titan. Strong electric charge.

Note that the orbit is stable in a 3:4 relationship with the larger moon Titan. So its position in space is well-predicted, but the orientation is not.


You can achieve what you want with Axial Precession. This actually happens here on Earth, although at a very slow rate (26,000 years for one rotation, this is believed to be a contributing factor to ice ages here).

If your planet or moon had a faster, more erratic axial precession (say on the rate of 1 or 2 per year or 1 every 2 years, with a large tilt angle), or even better - multiple (3 or more) axis of rotation you could have highly variable day/night lengths. This would lead to some very erratic season as well.


Something like that, though less extreme, could be possible with a fairly eccentric orbit.

Let's say the eccentricity is 1/3, which means that aphelion (the farthest distance from the sun) is twice the perihelion (closest distance). From Kepler's Laws we then have that the orbital speed at perihelion is twice that at aphelion.

As an approximation, let us assume that the orbital velocity is equal to the perihelion velocity for the entire quarter orbit nearest the sun (which is fairly close to the truth) and let us say that this quarter orbit corresponds to the longest day, from sunrise to sunset, or 48 hours. The planet then must make three-quarters of a full rotation during this quarter orbit, meaning that the rotation period of the planet is 64 hours.

At the aphelion, the planet would need roughly 96 hours to complete a quarter orbit (again as a close first approximation). This is one and a half rotation corresponding to to one and a quarter day/night cycle. A day is hence appromimately 38.4 hours.

This is less extreme than what you ask for, but a more eccentric orbit would imply even more extreme temperature differences than here, where the planet receives one-quarter the sunlight at aphelion than at perihelion. A very thick atmosphere could comnpensate for this, but not much more.

Another thing is that the total orbital period of the planet here is only 288 hours, or 12 Earth days. This implies a very cold sun, most likely a red dwarf. If you don't want that, this is not the solution for you.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.