As the Sun's luminosity slowly rises, the Earth's surface temperature will climb. Will Earth ever be too warm to have any more glacial periods? If so, when will that be?

Edit: The existing answer misunderstood the question. I'd like some boundary conditions, e.g. "when the sun's luminosity is 10% higher and the Earth becomes a 'moist greenhouse'" and "when the Sun turns Earth into a cinder". Those conditions would imply that the last Ice Age will happen 1-4 billion years from now. A better estimate would be awesome. Unless there's something I'm missing, with Stefan-Boltzmann and some reasonable assumptions (e.g. linear increase in luminosity) it should be possible to get a back-of-the-envelope calculation for this.


4 Answers 4


At the moment, nobody knows for sure whether or not changes in the Sun's luminosity can cause long-term climate changes. There have been various studies done, most since the advent of global warming, that have tried to address this. The conclusions have been interesting:

  1. The Sun is not responsible for the recent warming of the Earth.
  2. It's not yet known whether or not the Sun has influenced climate in the past.

Here are some studies (one supporting each side of the long-term debate):

  • A report by the National Research Council (available here) found that sunspot activity could be a major contributor to small changes in the Sun's brightness. There is a periodic trend in sunspot activity that lasts 11 years, and studies have considered this when looking at long-term trends. One target of inquiry is the Maunder Minimum, a period of low sunspot activity:

    The Maunder Minimum coincided with the Little Ice Age, a short period of low temperatures. In fact, the Sun may be entering a similar period within the next few decades. Observations could show whether or not there is a link between the two events.

    However, the NRC concluded that it is hard, if not impossible, to find any data supporting theories that there is a long-term patter. Hal Maring summarized it best:

    Hal Maring, a climate scientist at NASA headquarters who has studied the report, notes that “lots of interesting possibilities were suggested by the panelists. However, few, if any, have been quantified to the point that we can definitively assess their impact on climate.”

  • In the paper Possible solar origin of the 1,470-year glacial climate cycle demonstrated in a coupled model by Holger et al., periods called Dansgaard– Oeschger events were studied. These are climate oscillations lasting 1,470 years. Holger et al. analyzed ice cores from Greenland and attempted to model solar forcing, the amount of energy absorbed by Earth or radiated into space. They also used the ocean and atmosphere to account for changes (the NRC report, too, concluded that solar effects can trickle down through the atmosphere and other parts of Earth). Their conclusion was that, despite the fact that no records of 1,470-year solar cycle exist, the analysis supported a connection between solar variability and changes in Earth's climate.

    Before anyone gets too excited, I'll add this: 1) I haven't been able to find any comments from other scientists on the paper, so I'm not sure how well-received it was, and 2) The 1,470-year cycle can not account for long-term glacial periods, or the recent rise in global temperatures.

Another potentially interesting (and paywalled) paper can be found here. An informative website is here, but I'm not sure if it's entirely neutral. It does seem to present both sides, though.

I'll end on this: Whether or not the Sun severely impacts the Earth's climate, Milankovitch cycles are the biggest influence on ice ages.

It will take some pretty drastic changes in the Sun's luminosity to change the glacial periods . . . which the Sun will be undergoing in about billion years or so. It will still be on the main sequence, and will stay there for 4.4 billion years more, but it will have brightened by about 10%.

Once of the consequences of this will be that the Earth will lose its oceans. Through a process called the "moist greenhouse" effect, water heats up and reaches the atmosphere, where the hydrogen atoms are freed from the oxygen atoms and escape to space. This occurs at a tiny, tiny rate at present, but in about one billion years, the Sun will be bright enough that this will occur at a catastrophic rate.

That places the last glacial period somewhere between now and one billion years in the future. To figure this out, we have to look at long term patterns, such as Milankovitch Cycles. This graph illustrates them best:

From the present day to the far right, though, is only 800,000 years. That's unhelpful. This graph, while a bit clearer, shows only the variations without accounting for any changes in the Sun's luminosity (though that will not affect us during the next 800,000 years!).

Here's the actual answer to your question: We don't know. That's because some parts of the Earth will be affected more by the Sun's brightening that others. For example, the NRC found that the Pacific Ocean experienced a strong relationship with solar activity. So as the Sun gets brighter, some areas could still be strongly affected by Milankovitch cycles, while others might find themselves getting hotter than they would normally be.

We can be sure of one thing, though: It will certainly be before one billion years from now!

  • $\begingroup$ "It will take some pretty drastic changes in the Sun's luminosity to change the glacial periods . . . which the Sun will be undergoing in about billion years or so." This is exactly what I am referring to. Is there any way to get more specific? What bounds can we set? $\endgroup$
    – Kaia Leahy
    Feb 15, 2015 at 9:26
  • $\begingroup$ Where are we on that 11 year cycle? Are we headed for some nasty solar activity or a calm period? $\endgroup$
    – Mazura
    Feb 15, 2015 at 11:15
  • $\begingroup$ @Mazura We're at the lowest point - a calm period. $\endgroup$
    – HDE 226868
    Feb 15, 2015 at 14:08
  • $\begingroup$ @sethrin I'll edit it in to my answer when I can, but we can use the formula for effective temperature to give a really rough estimate for when average temperatures will have risen enough such that ice can no longer exist in large quantities. But we also have to account for human activity warming the Earth. If we stop soon, the increase over the next few centuries won't be catastrophic. If we keep going at our current rate, the glacial periods will be severely impacted. $\endgroup$
    – HDE 226868
    Feb 15, 2015 at 14:18
  • 2
    $\begingroup$ You're asking me to figure out when the last glacial period will be. The gist of the answer is this: We don't know. We don't know how much human activities are going to impact the environment. I think you underestimate the drastic changes that can occur if we keep going at the rate we're going. We don't know what temperatures will be like in 1,000 years, let alone 500,000,000 or 1,000,000,000. $\endgroup$
    – HDE 226868
    Feb 15, 2015 at 20:06

In my opinion, there's 3 primary factors. There's a difference between ice ages and ice age periods. The Milankovich cycles appear to play a key role in the forming and receding of individual ice ages, but what it doesn't appear to do is trigger ice age periods.

The modern Quaternary ice age period began about 2.58 million years ago. Milankovich cycles likely began long before then, so it's unlikely that Milankovich cycles triggered the period, only that they play a role in the cycle within the period.

Same is likely true for solar maximums and minimums. They come and go, but they aren't likely the drivers of ice age periods.

Timeline of glaciation

the Quaternary, the Karoo and the Andean-Saharan ice ages all happened in the last 450 million years, and they were separated by long periods so the cause should be looked at more long term. Milankovich cycles operate on 26,000, 42,000 and 100,000 timelines, much to short to drive changes over millions of years.

OK - the 3 things.

Land and ocean placement, Solar output (long term, not short term sunspot changes), and CO2

Land and Ocean Placement

One thing that seems necessary for ice age periods is a large land mass over one or both of the poles. The reason for this is pretty straight forward, land gives ice a place to grow and as land ice grows into a glacier, Albedo increases and oceans drop. High altitude over the glacier tends to blow cold air and maintain the glacier. Also as oceans cool, oceans can hold more CO2, dropping atmospheric CO2 and reducing the greenhouse effect. These feedback loops tend to make ice ages colder and they are only possible with land over one or both of the poles. Antarctica drifted over the South Pole and began forming a glacier about 30 million years ago and, in maybe 15-25 million years, it will drift off the the south pole and begin to melt. I think, once Antarctica drifts into warmer waters, ice ages are likely to stop. 15 million years or so? Maybe more?

It's also possible that Canada or Russia drift over the North Pole and as the the giant Antarctic Glacier is replaced by a giant Arctic Glacier, but I think the general consensus is that we can expect ice ages to stop when antarctic drifts over warmer water. We can also expect higher sea level too, lower Albedo and no more ice age cycle.

Solar output - long term, not short term sunspot cycles.

The sun growing more luminous, meaning, more energy for Earth, it's expected to grow warmer by about 10% over the next billion years - roughly 1% every 100 million years. (Source). 1% might not sound like much but it actually is. Solar variation, which is believed to have caused the Mini ice age, has a variation of about 1 - 1.5 watts per square meter. (Source)

A 1% more luminous sun would be in the range of 13-14 watts per square meter, so as the sun grows brighter, and if everything else stays the same, certainly within 100-200 million years, the sun would be sufficiently hot to make ice ages pretty darn unlikely. This site suggests that a 10% increase in solar luminosity could result in 47 °C, 116 °F. It will take quite a bit less than that to end ice ages. (Source)

We also have to consider orbital drift. If the Earth is slowly moving away from the sun, that's a factor too, and by this study, it is moving away, at 15 CM per year (Source), so, in 100 million years, 150,000 km, 0.1% further from the sun, about 0.2% less solar irradiance, which is only 20% of the increase in solar output,so this, at least by current estimates, will be overshadowed by the sun growing brighter. (Source)

Carbon Sequestering / atmospheric CO2 PPM

Finally, and very difficult to predict over the long term is carbon sequestering. CO2 has generally been decreasing over millions of years and the drop in CO2 may have played a role in the current ice age cycle. It's entirely possible, though some might disagree, but it's certainly possible that our current uptick in CO2 by human activity has already knocked the next ice age off it's feet and we might have changed the planet enough to end the ice age cycle. The orbital forcing is believed to be only about 1 degree C but ice ages can be about 10 degrees C colder due to feedback mechanisms. Our man made increase in CO2 is expected to warm the earth more than that, so at current CO2 levels, it's entirely possible that an ice-age couldn't form. source


The uncertainty in the estimates from various climate models for doubling CO2-equivalent concentration is in the range of 1°C to 3°C with the probability distributions having long tails out to much larger temperature changes.

But, long term the CO2 isn't likely to stay at 400 PPM, though there's a good chance it won't drop to pre anthropomorphic levels of cycling between 180 and 280 PPM that we had during the recent ice age cycles. That might be enough to alter the recent ice age cycle, though long term predictions about CO2 sequestering are very difficult, but I do personally believe that by burning fossil fuels, we may have actually put an end to the ice age cycle. I'm by no means stating that as a certainty, but it's within the realm of possibility.

We've also seen CO2 levels gradually drop over millions of years, so it's possible, in a million or two years, we'll see CO2 levels even lower than we've seen and perhaps a resurgence of ice ages. Continental drift over the next few million years is predictable and solar output over longer periods is somewhat predictable, but CO2 sequestering is enormously difficult to predict - so that's the wild card in this estimate.

enter image description here

As we can see from the chart, there's been a pretty steady drop in CO2 PPM over the last 15 million years or so. As the ocean grows colder it can store more CO2 so there may be a correlation between falling temperature and falling CO2 PPM - one of the primary climate feedback mechanisms, but predicting future CO2 levels - I wouldn't even know what to guess. Certainly low enough CO2 PPM could lead to further ice ages, even after Antarctica drifts off the south pole, so, there's the problem - too many unknowns. This question has no answer.

A 4th possible key driver would be ocean currents, as indicated by the theory that the formation of the Isthmus of Panama may have been a key driver in the current ice age cycle, but predicting ocean current changes millions of years into the future is above my pay grade.

The 3 primary drivers do seem to fit past ice ages.

The the Andean-Saharan ice age period 425-450 million years ago, CO2 levels were much higher but that long ago the sun was perhaps 3%-4% less luminous and much of the land was over the antarctic circle.

enter image description here

The Karoo ice age period is associated with a drop in CO2 and an increase in O2 (Source)

That probably long and rambling and needs more than a little clean-up and I invite counterpoints, but I think that's the gist of estimating when ice ages will end. A precise answer isn't possible.

  • $\begingroup$ A precise answer isn't possible, but it is possible to set bounds. This is considering the right question and I believe the right factors. If we were to summarize, we could say "somewhere between 15 million and 150 million years from now". Thanks for answering. $\endgroup$
    – Kaia Leahy
    Aug 17, 2015 at 2:38

It is very hard to exactly tell when the next ice age will happen, to tell about the last is even harder.

If we support ourselves in the present state of knowledge, the Sun is in the middle of its life span, the star evolution theory (see: https://en.wikipedia.org/wiki/Stellar_evolution), which means that at certain time it will start to cool down and in the next cycle it will become a giant red, increasing its volume until it engulfs Earth, destroying it. It will grow to reach a radius in between Earth and Mars.

I think that the last ice age will be very long, during the cooling and transformation in a Red start. It will finish with the destruction of Earth by the Giant Red Sun. It is expected to happen around 4.5 billions years from now.


I believe the final ice age will happen around 100 to 150 million years from now and the sun gets 1% hotter every 100 millions years and after about 100 million years or so it will be too hot for any ice ages to happen after that point in time.

  • 2
    $\begingroup$ welcome to the site. You answer would be much better if you listed references for you statements. $\endgroup$
    – Fred
    Jul 31, 2015 at 15:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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