If we have extreme and short winter caused by 3-day-eclipse, what would be temperatures like?

Planet is like earth, atmosphere, oceans, ice caps and so, but there is no regular seasons, because axis is made nicely without moon to shake it much.

Double star system (one light and one dark). When stars are in straight line, there is short winter in planet, lets say 3 days (even if its physically though to get such a long eclipse).

So we have 3 days with cca. 1% of sun light. After that, there is summer again and there is no "other" winter in the planet, so temperatures could come back in few days.

What would be temperatures in day 1, 2 and 3?
What would be sea level in day 1, 2 and 3?
What will be the direction and speed of wind?

My assumptions are this (please do not agree with it :-) ):
Day 1: 10 to 5 degrees - First day of winter is mild, just like longer night
Day 2: 0 to -20
Day 3: -40 to -70 (?)
Could be this fair enough?

Sea levels - because of fact, that volume of ice caps will be much greater till first day of winter, and they will consume lot of ocean water, sea level should be lower from day 1 with strong currents to poles.
Day 1: -0,5m
Day 2: -1m (world ocean is completely frozen to the few metres of depth)
Day 3: -1m (in coastline where all water is frozen and have no space to go lower) to max -3m in open waters

Thanks for clarify and your numbers. Miki

  • $\begingroup$ How long is your year? Three days without sunlight is nothing -- you could compare that to any decent stalled hurricane on Earth. $\endgroup$ Jul 29, 2020 at 17:27
  • 1
    $\begingroup$ Where do you live? In most places of the world it is pretty common to get three days in a row with thick clouds. Nothing spectactular happens. (And the idea that the ocean would freeze over in a day is straight out of a bad Hollywood movie. No, it would not freeze completely to a few meters of depth. It wouldn't even get a thin layer of ice. The latent heat of solidification of water is huge.) $\endgroup$
    – AlexP
    Jul 29, 2020 at 21:11
  • $\begingroup$ Thanks for this view guys, better work needed in the book for readers to understand. As for the comparison huge clouds vs. eclipse - there is significant difference. During the huricane or just huge clowds, there is still lot of sunlight that goes to the atmosphere, so there is lot of neighbour areas, which can make the ground hotter (not so much, but quite). So even if some area will get cooler by 5 degrees, there are still other sources of hot temperature like 10 km away, which can constantly bring heat. $\endgroup$
    – miki
    Aug 1, 2020 at 12:29

1 Answer 1


Not much, but it would be noticeable. Averaged over a year, Earth's surface receives 240W per square meter of sunlight source. The Earth's surface temperature, on average, is actually quite constant. We don't constantly have ice ages and runaway greenhouse effects. Ignoring the effects of the season, our average temperature is 15C. This is, crucially, because the amount of energy in equals the amount of energy out. Earth is, in short, in thermal equilibrium with the space environment it lives in.

So we're starting at 15C average global temperature. I can't do this calculation globally, so instead let's do it locally and estimate. Say at my location it is 15C and a calm, clear day. At night, because it is very calm and the atmosphere kind of sucks as a thermal conductor unless it's windy, the major cooling effect is due to radiative heat loss (well, and the upper atmosphere doing its thing, but it's always cold there so let's not worry about it).

According to a very helpful blog, at these approximate temperatures air temperature will drop by about 1 degree C per hour in the night (see the graph) but will slow down as the temperature goes down. In fact, by that graph, the rate at which temperature decreases slows to about half a degree per hour by midnight.

So, time T = 0 hours the eclipse starts. T = five hours or so and you're at 10C. It will likely take until the next day until you're at 0C (T = 24 hours). After that, expect a decreased rate of cooling, but probably nothing horribly slow. Ignoring that, it looks like you can expect around -20C by the end of the third day. Probably higher (that is, closer to zero) because the Earth is a giant heat sink and it will release a lot of heat into the air as it cools through radiation.

So it would be time for a coat, but nothing catastrophic.

The poles would experience much colder temperatures than the equator, but they're already doing that anyway.

Your oceans would not have NEARLY enough time to do anything. They are HUGE, HUGE heat sinks. I'd be surprised if large ponds started freezing over.

  • $\begingroup$ Hi, really thanks for this summary (and link which basicaly proofs it - I am especially sad about slow progress of antarctic temps, that will fall down to -60 after Months without sun - but there is another half of planet to heat it at least a bit). Anyway, lot of things to agree and lot of things to think about to build this world properly, so huge thumbs up. I hope I will find sort of conditions that will support much extreme temperatures (which have to be there due to more interesting story). $\endgroup$
    – miki
    Aug 1, 2020 at 12:51
  • $\begingroup$ 1) If it is thinner atmosphere in my planet (with high speed winds there), the heat will go away much faster to the outer space 2) If local sun is further, there should be constant lower tempreatures on whole planet, maximum summer temp of 15 or 20 deg should get me closer to winter extremes (hopefully). So if it is just a bad rainy week of 5 to 10 deg right before the eclipse, we are even close. 3) Lower co2 volumes will support anti greenhouse effect, so heat will cumulate slower in atmosphere. $\endgroup$
    – miki
    Aug 1, 2020 at 12:51
  • $\begingroup$ I suppose it depends on what you need in your story. You already have a "dark Star" and such objects do not exist, so you are a bit more free to come up with a reason that it gets colder. Perhaps the planet is in fact a giant artefact that absorbs energy from sunlight and heat and during the eclipse it feeds off it's atmospheric heat to stay alive. Something like that would do what you need $\endgroup$ Aug 2, 2020 at 2:31

You must log in to answer this question.

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