Here on earth, it takes about 90,000 years for major ice sheets to build up, and another 10,000 for them to collapse, resulting in 100,000 year cycles of major glaciation, from what I understand.

Would it be too far fetched to imagine a world where this happens at a much faster and more regular rate, e.g. every thousand years? Could my explanation for this phenomenon be as simple as a periodic gravitational tug from another large planet? And what about extreme glaciation, where the ice sheets extend so far from the poles that only the equator remains uncovered? Would there be enough time for that to happen?

I understand that ice sheets form when summer temperatures don’t rise above freezing (due to orbital eccentricities, changes in ocean currents, etc.) and snow continues to accumulate and compress year-round. But I don’t have any sense of how long my scenario would take to unfurl.

Thanks in advance for any pointers!

  • 1
    $\begingroup$ Please do the research. Glaciation happens on 100,000 year timescales, with 20,000 year interglacials or times like ours. Glaciation happening every thousand years in extremely short term. Glaciation may happen quickly, possibly within millennia. You can start here: en.wikipedia.org/wiki/Timeline_of_glaciation $\endgroup$
    – a4android
    Nov 22, 2019 at 6:47
  • 1
    $\begingroup$ Careful with those "major glaciations" which occur on a hundred million year timescale: the ice sheets advance and retreat during those glacial periods. In fact, we are in the middle of one right now! What you are looking for are glacial maxima within a glacial period, and those happen on much shorter timescales (about one hundred thousand years between them). $\endgroup$
    – AlexP
    Nov 22, 2019 at 12:41

3 Answers 3


Younger Dryas was an abrupt event with roughly the feature you require, except the periodicity.

The Younger Dryas (around 12,900 to 11,700 years BP) was a return to glacial conditions after the Late Glacial Interstadial, which temporarily reversed the gradual climatic warming after the Last Glacial Maximum (LGM) started receding around 20,000 BP.

The change to glacial conditions at the onset of the Younger Dryas in the higher latitudes of the Northern Hemisphere, between 12,900 and 11,500 calendar years BP, has been argued to have been quite abrupt. It is in sharp contrast to the warming of the preceding Older Dryas interstadial. Its end has been inferred to have occurred over a period of a decade or so,[14] but the onset may have even been faster. Thermally fractionated nitrogen and argon isotope data from Greenland ice core GISP2 indicate that its summit was around 15 °C (27 °F) colder during the Younger Dryas than today.

The reasons behind it are debated, there is the impact hypothesis, the eruption hypothesis and the supernova hypothesis. Probably the supernova is the least plausible in term of periodic recurrence at the right distance, but the other two might be plausibly periodic.


One thousand years is not long enough to build a massive ice sheet. But it might be achieved in several thousand years.

Oceanic currents take around 1000 years to cycle so I doubt 1000 years is sufficient time for "the Earth to freeze up totally except for the equatorial regions". Warm water would still be welling up from below.

What might help would be a very large and constant air current that moved damp air continuously over the Polar Regions dumping massive quantities of snow and a suitably shaped land mass to allow continued build up. It would help but 1000 years would not be enough time.

If 20cm of snow fell every single day (roughly compacts to 2cm of solid ice) it’s only equivalent to 730m of ice in 1000 years (assuming no melt or run off) which is probably not enough. And such a massive continuous snow fall is not realistic anyway.


Could my explanation for this phenomenon be as simple as a periodic gravitational tug from another large planet?

This is known as an exchange orbit, and occurs with Saturn's moons of Epimetheus and Janus. Two co-orbiting bodies with similar mass and orbital characteristics can exchange eccentricity and/or orbital distance during close passes. Orbiting further from the primary or on a more elliptical path would definitely reduce solar flux and therefore surface temperature.

There are two problems, though. One is that, as explained in Slarty's answer, this isn't enough time for proper glaciation (although that's not much comfort to anyone living through it).

The other is the orbital mechanics required. Co-orbiting planet-sized are probably not stable over the long term, especially at a significant distance from their primary (i.e., in the habitable zone). There's also the question of how two bodies formed there without merging into one body. (I suspect putting the right numbers into a Theia-type impact would end up with two co-orbiting bodies for awhile, but I don't know the math on that one; I doubt it'd be stable.)

Also, the Epimetheus-Janus exchange orbit takes only four years to run through a complete exchange, from the point where one moon enters the outer track to the point where it leaves the inner track again. I suspect this is tied to its short orbital period (less than a day) but I don't know how it scales for planets' orbital distances and periods. My instinct is that the longer the time between conjunctions, the more susceptible the resonance will be to being disturbed by another body.

TL;DR: Yes, it's possible to have a co-orbiting body raise and lower your planet's temperature but that doesn't necessarily mean glaciers and it's suspect how stable it is over the long term.


You must log in to answer this question.

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