I am trying to create a planet for a fantasy role-playing game where the equator is an impassible region of ice. Ideally, I want to find a science-based solution to this and not have to resort to explaining it away with magic. Here's what I am trying to create.

  1. Planet has seasons. (If there are ways that seasons can be created without giving a planet tilt, I am open to this.)
  2. Planet is cold at the equator.
  3. Planet is hot at both poles.
  4. Planet has a day and night cycle similar to that of Earth.
  5. Planet can have any year length. Ie. If the planet has a longer orbital period due to a highly elliptical orbit.

Things I have thought of so far:

  1. Give the planet a 90° tilt. This would cause half of the planet to be in constant darkness/light in violation of Criteria 4.
  2. Give the planet 0° tilt and rings that block out a good portion of the sunlight about the equator. With 0° tilt, the planet wouldn't have seasons (breaking Criteria 1) unless there is some other way of bringing them about. In addition, most rings are basically flat and the shadow cast by rings at 0° tilt about the equator would be very narrow.

Thanks all!

  • $\begingroup$ My solution would be to hang the planet in between two binary stars at the point where each star's gravity is equal to the other. Unfortunately, I postulated such a static solar system in an earlier question and was told that such a system would be unstable. $\endgroup$ Mar 8, 2017 at 16:31
  • $\begingroup$ @HenryTaylor Do you mean something similar to a Sitnikov Planet? I am okay with a system that has an unstable or highly improbable orbit however, I haven't been able to wrap my head around what a day would be like on a planet in such a binary system. Any thoughts on what a day/season/year would be like on a planet in such a system? Thanks again! $\endgroup$
    – TGratton
    Mar 8, 2017 at 16:38
  • 2
    $\begingroup$ Possible duplicate of Planet with poles warmer than equator $\endgroup$
    – L.Dutch
    Mar 8, 2017 at 17:18
  • $\begingroup$ Not a duplicate due to different constraints. $\endgroup$
    – kingledion
    Mar 8, 2017 at 17:43
  • $\begingroup$ Give it an elliptical orbit, and you have seasons. $\endgroup$ Mar 8, 2017 at 18:25

6 Answers 6


Lets start with (4). To have day/night cycles the planet must rotate and there must be a local sun (or suns) of some sort typically at less than 90 degrees from the planet's rotation. Check.

Add (2). With the sun overhead(ish) at the equator, in order to be cold there the day/night sun needs to far enough away that its average heating during a day is less than whatever causes (3). There are no orbital mechanics that permit a body to remain over the poles more than the equator all year round. This will make the world darker but keeps day/night effects.

Now add (3). In order to be hot at the poles, there needs to be a heat source. Lets try adding a brown dwarf. They give off heat, but not a lot of light. If we put one over the pole however, our planet would have to orbit the brown dwarf. After a quarter orbit, the brown dwarf is now over the equator. (This is the same problem with your suggestion 1 above. A tilt of 90 degrees only has the sun over the pole at one point in the orbit. This is more like extreme seasons rather than being tidally locked.)

So to get the heat at the poles you need a source that is internal to the planet. One idea is to use a heated core with magma convection currents that raise up in the poles and sink at the equator. The poles would then get more heat from the core. But this is the reverse of what the natural flow would be due to the planets rotation. You would need most of the magma to be a material that got MORE dense as the temperature increases. There are some examples of this, but not many. Also you would need the core to be extra hot, possibly due to radioactivity.

In this situation weak seasons can be had like on earth from the, here distant, sun.

Suppose we dropped the whole year round idea, and only need this situation to be stable during one season. Then we could have the planet co-orbit a binary star system at the L4 or L5 Lagrange points with a tilt of 90 degrees. Then every other season one star would shine mostly on one hemisphere or the other, leading to polar heating more than the equator. Unfortunately this means that in the other two seasons one of your two poles are plunged into never ending darkness.

Lastly we return to your excellent idea of having a wide set of rings block the equatorial light on a low tilt planet. All you need now is seasons. But tilt is not the only way to have seasons. Instead you can make your planet orbit one of a pair of stars in a binary star system. The close star provides most of the heat, but the distant star provides variability in the small heat it adds by where it is in its orbit. (This is what the planets around alpha centari A experience from alpha centari B.) Plus by changing the inclination of star B you can decide if the northern and southern hemispheres experience the same seasons at the same time or alternate.(Or make your planet a binary with a brown dwarf on an elliptical orbit. Or just make the planet's orbit elliptical itself for short summers and long winters felt by both north and south hemispheres at the same time.)

TL;DR: Use your rings idea but add a distant binary star for seasons.


Yes, it is possible, although the flavor would probably be wrong for what you have in mind.

Climate varies with latitude and altitude. Mountains on the equator have permanent glaciers.

So to have the equator closed by ice you just have to make it high enough.The easiest way would probably be to make the planet less spherical. This has the benefit of not needing mountains, although you'd need to keep the equator free of of oceans, ie. have a supercontinent that encircles the planet on the equator.

This state is not stable as planets are by definition in hydrostatic equilibrium. So you'd need some reason, such as divine or alien interference that explains why the planet is less spherical than it should. Planet had too fast rotation and the gods fixed it, maybe?

Alternately, since you need the supercontinent anyway, you can just fiat that every point on the equator just happens to be high enough thanks to various highlands and mountain ranges. There is no real reason why not. Note that since the equator would have much less heat absorbing ocean and much more heat reflecting ice and snow, it would not have to be as high as is in our world for permanent glaciers. There would be permanent winds away from the glaciers and without ocean currents, there would really be no way for heat from surrounding warmer lands to melt the equator.

Making the poles hot is more problematic. With huge glaciers and no seas on the equator it is perfectly reasonable to assume mostly dried out ocean basins surrounding the poles. If you configured what remains of the seas for efficient heat transport, I'd guess there is nothing stopping the poles from being hot.

So you'd have a polar ocean with water heated closer to the equator. There would be islands and coastal lands with geology largely formed of marine evaporites.

The problem with this solution is that it would do nothing about the poles being short on sunlight at least half the year. So the agricultural productivity and supported population levels would not be that good.

Then again, I doubt it is that crucial to you to have huge farmlands on the poles? Since you need polar oceans anyway for heat transport, you can place only islands with hardy fishermen on the poles and have the agricultural lands on the coasts closer to the equator with better sunlight.


In a two-star system, there is no stable planetary orbit near the stars. Planets will

  • most probably go away (and then they will either leave the stellar system, or get a stable orbit far from the stars)
  • get to an orbit very close to one of the stars (and thus, only its gravity will have a significant effect on them)

There is no third possibility. If you want a stable planetary orbit, only a single star can affect their orbit significantly.

Some tricky solutions (for example, a trajectory forming an "8") are all very unstable, and the planet will leave them fast.

The only exception is if the L4 or L5 Lagrange-point of a two-star system. But in this case, you can't have the required constellation of the Suns.

I can imagine 2 possibilities for a similar planet.

  1. Consider a "planet" (more exactly, red dwarf star) orbiting a Sun-sized star around in the current Sun-Mars distance. An Earth-sized "moon" of it would be tidal locked to it. Thus, it will be a constant "Sun" over its North pole. The "second Sun", on the place of our Sun, would be able to heat it mainly from the southern direction. The result will be a hot northern hemisphere with a constant Sun, and a colder southern hemisphere with a moving, secondary Sun. If you set the orbital parameters correctly, maybe you can get a cold equator.
  2. You are in a single-star system, with a tidally locked Earth to the Sun. This Earth should be around between the Mars and Jupiter distance from the Sun. This results a relatively warm northern hemisphere and a very ice southern one. Some gods or interstellar ancient civilization could have built a large, stable space mirror in the L2 Lagrange points of its orbits with the goal to stabilize its weather. Although the L2 lagrange point is unstable for point-like masses, it may be stable for a large mirror. If the mirror is focused only to a small part of the ice southern hemisphere, you will get the warm poles and the ice equator.

In both cases, the atmosphere of this planet should be rarer as ours, to avoid the heat transfer (for example, a purely 20% oxygen atmosphere, without our nitrogen would be okay).

If you wish I can give some pictures from the actually needed constellations.


Don't overthink. There is no need for some elaborate exotic orbital mechanics.

  1. Seasons: Axial tilt, just like Earth. If you want to avoid Earth-like seasonal patterns, the answer here entirely depends upon exactly what kind of seasons you want. For something erratic (or an unusual pattern), use fluctuations in solar cycles changing the output of the star, or if you want both poles to have the same seasons at the same time then use no axial tilt and vary temperature by a highly ecliptic orbit.

  2. Cold equator: High elevation at equator due to a more oblate spheroid shape of the planet makes crossing it like climbing Mt Everest. This eliminates the prospect of frozen seas crossing between the poles, but keeps the hemispheres separated by frozen peaks and glaciers.

3: Hot Poles: Hot planet. Strong greenhouse effect, possibly backed up by volcanic activity, keeps the poles toasty warm. Thermal hot springs would be common if using volcanic heat. If you're doing high-tech sci-fi scenario, this may be relevant, but otherwise you don't need to explain the details of the meteorology to pre-scientific characters.

  1. Day/Night like Earth = spins like Earth. Simple.

Just how deep into the details do you need to get in explaining and plotting out the orbital mechanics of every astronomical body in your world? Don't overthink and don't over-explain - is your target audience really that interested in getting into the details of examining the realistic charting of magma flows in the mantle of your planet?


One possibility would be to give the planet some sort of long-term cloud system forming a belt around around the equator that blocks out sunlight, or perhaps even having the clouds form high-altitude ice crystals that reflect sunlight away. You could still have day-night cycles even at the equator, it is just that the equator blocks enough of the light to reduce temperatures.

Some of the gas giant planets have storms lasting for centuries or even thousands of years that form bands around the planet vaguely similar to what I am describing. However, I am not sure what specific set of geographic features would encourage the formation of such a weather system on a rocky planet.


The procession of Mercury fits some of your criteria (all but #4), in spirit if not to the letter. Each year as Mercury's orbit brings it to perihelion, it "locks" one face in day/summer. The next time Mercury comes in close it locks the other side making night/winter. The opposite faces of Mercury alternate in 3:2 resonance with Mercury's year, with a twilight ring that wraps up and down over the poles rather than around the equator. It wouldn't be a perfect twilight zone, each twilight face would see the sun at aphelion every other year.

enter image description here


You must log in to answer this question.

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