I can't decide if I should use a habitable moon orbiting a giant, or a planet of similar or larger size between my planet and its sun.

Anyways, it'll go through a long eclipse each year, lasting for about a month.

And wouldn't "sister" moons around a giant prolong an eclipse every few years? Or rather... could they?

I need my eclipses to cause a long winter season with total darkness, and if possible maybe an extra "extra long" one every now and then.

A very important plant depends on that.

the system

On this pic my home-world is the second planet.

Possible variant: the second planet, my home-world, would be a moon of the third and this one would take the second orbit.


  • $\begingroup$ Nothing can move in a way to close an eclipse for the whole year, unless it is stuck in L1 Lagrangian point $\endgroup$
    – Alexander
    Sep 20, 2017 at 0:01
  • $\begingroup$ Sorry. I meant for a month in the year, more or less. And my home planet would either be the moon around the giant or a plain earth-like planet. Thanks $\endgroup$ Sep 20, 2017 at 0:55
  • $\begingroup$ I have added the information about what this planet is and the length of the eclipse to your question, and tried to clarify things based off of your comment. You can edit your question using the 'edit' button. Generally, we like seeing updates in the question itself rather than in the comments- it makes things easier for people to read and presents all the information up front. I think this is a good question, and would like to see what people come up with. $\endgroup$
    – PipperChip
    Sep 20, 2017 at 1:29
  • $\begingroup$ I can only think of Saturn and it's planetary rings(let's call them nano-moons) then depends on location you can have up to thousands of years of eclipse! $\endgroup$
    – user6760
    Sep 20, 2017 at 3:17
  • $\begingroup$ Need far more than 15 minutes long eclipse... complete and total darkness and happening only once each year. $\endgroup$ Sep 20, 2017 at 6:42

2 Answers 2


The answer to this question lies in three factors:

  • The orbital period (one year) for your planet in question
  • The rotational period (one day) for your planet in question
  • The orbital period of the body causing the eclipse on the planet

The easiest way to accomplish an eclipse of this type would be (as you suggested) a habitable moon orbiting a larger body (think Pandora from Avatar).

If your planet orbits its parent in the same time period as the parent orbits the star then depending on how the orbits work out you end up with ~1/4 of the year being in an eclipse. This also lets you maintain a normal day/night cycle on the planet without going into shadow of its parent.

Diagram added for clarity:

enter image description here

  • $\begingroup$ Thanks. That's how imagined it. And wouldn't other (sister) moons cause longer eclipses as well? $\endgroup$ Sep 20, 2017 at 2:49
  • $\begingroup$ They very well could. Orbits are very diverse and how they align with each other can cause many, many different effects. Just in our own solar system we have a planet with 60+ moons, a planet that rotates 90 degrees sideways, a planet with a day longer than a year, and a planetoid that crosses another planets orbit $\endgroup$
    – enpaul
    Sep 20, 2017 at 3:11
  • $\begingroup$ Provided the moon has a circular orbit and orbits in the plane of the ecliptic this might work. The difficulty is that for this to work the orbiting body needs to be big and/or close and slow. One way to slow it down is to reduce the mass of the body it orbits but detailed calculations would be needed to see if it was possible as you can’t just slow an orbit down without increasing the orbital radius. $\endgroup$
    – Slarty
    Sep 20, 2017 at 8:04
  • $\begingroup$ The astronomical math would not work the way the diagram suggests. Shadow cone from the giant planet would be much less than 25% of its moon's orbital path. $\endgroup$
    – Alexander
    Sep 20, 2017 at 22:18
  • $\begingroup$ @Alexander yes that's correct, though I thought the approximation would be sufficient here. The OP was focusing on the orbital mechanics of arranging a multi-month eclipse. The exact calculations will of course work out differently. $\endgroup$
    – enpaul
    Sep 21, 2017 at 1:30

I wonder if you would like to have the Eclipses take place at irregular intervals? if so - having a regular Earth like planet that circles the primary on it's own BUT another much larger planet circling the primary inside the orbit of the Earth like planet could give you opportunities for Eclipses to happen at pretty odd intervals.

It wouldn't need to be a gas giant (And really couldn't be one either - such close proximity to the Sun would cause it to torch up and become a Binary system!) But it could be a large rocky body. OR just to make things really interesting - it could be artificial, perhaps even a giant hollow ball that has some kind of internal habitable zone... That's if you are going down some kind of SciFi road... Come to think of it - you could even have it be a NATURAL hollow sphere, there's plenty of people that have tried to explain why and how Pellucidor could exist for you to crib off some of that work...

Regardless of what you do on the celestial body causing the eclipse, have them be 2 separate planets in independant orbits is IMHO easier to design than a habitable moon. I've always felt a bit iffy about how an ecosystem can develop with such radical variations in the Day/ Night cycle a moon is likely to have. When you look at MOST of the moons in our own solar system they are tidally locked - this is going to make things complicated from a Day Night point of view.

The best way to try out what I'm describing is grab a Basketball (Or soccer ball) and a Softball and a Tennis ball. Put them in a line from biggest to smallest - the Basketball then becomes the sun, The Softball becomes the planet that creates the eclipse, the Tennis Ball becomes the planet your characters live on. You can then move things around a bit so you can see how far away from the tennis ball and the basketball the softball needs to be to fully eclipse things. You'll need a bit of overlap - so that your whole planet remains in shadow for a good long time.

My thinking was that the bigger planet would have a slower orbital period than the smaller one... BUT I took a quick google and found there are Kepler's laws of orbital mechanics that say otherwise - that in fact the further you go from the Sun the longer it takes to orbit the sun, Explanation here: https://www.exploratorium.edu/ronh/age/ SO for a bit of scientific consistency - you should go with the habitable planet orbiting slower than the one causing the eclipse.

That site also has some nice info about relative facts between the planets in our solar system that'd be quite useful for this whole idea...

Here is an image that demonstrates it working reasonably well (You will need to go to the site and copy the settings to see it working) https://i.stack.imgur.com/nmBVm.jpg

The Site is: https://phet.colorado.edu/sims/my-solar-system/my-solar-system_en.html

I played with the settings a fair bit before figuring out ones that gave approximately the results you were after - But these ones demonstrate you can probably achieve the exact sort of result you are after. Of course - you'd need someone smarter than me who specialises in celestial orbits to figure out what you will really need in terms of realtive mass & velocity.

  • $\begingroup$ Thank you. I experimented with a few simulators and kept getting two short eclipses each year and a rather irregular orbit for my earth-like planet. The inner larger planet still orbits a lot faster than my "home world". Lots to think about tonight... $\endgroup$ Sep 20, 2017 at 4:41
  • $\begingroup$ @shieldedtulip I added a couple of links above that demonstrate the result you are after (Couldn't get the damn thing to actually upload inline - I may come back and try to do that later) I did get it to have only one Eclipse a year with these settings - in fact it is one about every 1.25 years - but a bit of tweaking would get it to be about right. The orbits also are pretty stable and normal looking. $\endgroup$
    – kiltannen
    Sep 22, 2017 at 3:15
  • $\begingroup$ thanks, thank very much indeed. I actually used this simulator quite a bit and got it to sort of work with three planets, and by that I mean I managed to have 3 stable planets orbiting their star, but was still getting the eclipse twice per year. Have been playing with these simulations for years now... on and off trying to get the results I need. $\endgroup$ Sep 22, 2017 at 10:14
  • $\begingroup$ Actually... I did use this one but the one with 3 planets was this starchitect.net lost count of how many simulatorsI've used... $\endgroup$ Sep 22, 2017 at 10:27
  • $\begingroup$ Did you try the settings I suggested? I think they come pretty close to giving 1 eclipse per year - I didn't manage to figure out how long of an eclipse that would give - but it seemed like at the least it would be several days long... $\endgroup$
    – kiltannen
    Sep 25, 2017 at 2:31

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