Ice is strange, in that the solid form is less dense than the liquid water it freezes from. This causes Ice to float on top of the water.

How would our planet earth be different if something changed this property of Ice so that it became significantly denser than water. Dense enough to sink to the bottom.

Would the poles freeze up completely?

How would the climate and life be affected?

  • $\begingroup$ Kinda a grey area in your question...ice is most dense around 4 degrees...would solid ice being more dense than water also imply that 1 degree water would be more dense than 4 degree water? This would have some pretty massive implications on the heat transfer system on earth, possibly giving a very thin zone between an overtly warm equator and from poles thats actually usable for life. $\endgroup$
    – Twelfth
    Jan 5, 2015 at 19:28
  • $\begingroup$ @Twelfth Honestly I've not decided on the mechanism by which the density would change so that would be a possibility too. $\endgroup$
    – Tim B
    Jan 5, 2015 at 20:05
  • $\begingroup$ Add another one...liquid to ice (state change) takes quite a bit of energy to accomplish in the case of water. Would this same energy requirement exist if ice was heavier than water? Would their be an equivalent change in water to steam? Not sure if I can answer yet, just trying to flesh out some questions I'm running into. $\endgroup$
    – Twelfth
    Jan 5, 2015 at 20:21
  • 1
    $\begingroup$ I'd expect the state change energy to remain the same. That's present even in other materials with more conventional weight ratios. $\endgroup$
    – Tim B
    Jan 5, 2015 at 20:49
  • $\begingroup$ well to start fresh water life would not exist once you get away from the equator. $\endgroup$
    – John
    Feb 18, 2021 at 1:27

6 Answers 6


If ice did not float, Snowball Earth would have had a bigger effect on life ( when after great oxygenation it was not enough CO2 in the atmosphere to keep Earth warm enough and, according to some theories, the entire surface of Earth was covered by ice.) But it is quite possible that life could survive anyway around underwater volcanoes and in areas of increased salinity (saltier water freezes at lower temperature).

Maybe it was not 100% ice covered, maybe there were gaps (slushball earth) and after great extinction some lifeforms would have survived, but you are taking chances with lowering the chance of evolving more complex life forms. We know that Tardigrade can survive freezing and the vacuum of space, more complex life is separate question.

Even with today's climate, if ice sunk to the ocean bottom (where it is colder, and summer warmth is harder to get). Such ice would not melt for millennia, creating perma-ice (similar to permafrost). Life would be possible only in shallow layer of water between perma-ice and surface. This layer would be very cold, cooled from the bottom by perma-ice.

As a result, oil fields would not form, or if formed, they would be between layers of perma-ice. Same with other geological forms created from sediments of sea bottoms, like chalk, limestone, sandstone. These will be mixed with perma-ice. Possibly limestome would never form, because pressure to created it would heat and melt ice layers.

Such ice frozen under pressure transforms to hexagonal columns. Very pretty. Columns can be separated intact - ice shatters on impact into hexagonal columns.

After compressed in bigger depth, or uplifted by geological processes above sea level, ice would melt and layer above would collapse. Would create very different geology on such planet.

Interesting effects on such planet:

  • No tsunami after earthquake, because liquid ocean is shallow.
  • More volcanoes, because more water (perma-ice) would be subducted in subduction zones in continental drift.
  • Frequent smaller tsunami, when mixed layers of ice and sediment crack.
  • Bigger changes in weather between summer and winter, because temperature stabilizing effect of oceans would be substantially decreased (less volume of water).
  • Interesting effects with near-ocean volcanoes like Hawaii and Iceland. Or underwater volcanoes, melting perma-ice and messing up geologic layers, creating more tsunamis.
  • Warm ocean currents would melt underwater canyons in perma-ice. Would have interesting effects on oceans, especially as continental drift would move such canyons.
  • such canyons in perma-ice would have surprising ways to focus tsunami. After thinking about it, tsunamis on such planet would be frequent and more localized, more focused, because energy of tsunami would dissipate less (wave would travel down the canyon with little decrease in devastating power or dispersion).
  • These canyons would have similar effect like bends in river delta: instead of sediments, layers of ice higher than the "banks" of "canyons" would be deposited by tsunamis, and quickly frozen, creating "walls" between polders where water in summer would be fresh melt (little salt), but darker bottom because of sediments. Those polders would be freshwater lakes on perma-ice (not on dry land).
  • Liquid parts of oceans would by much saltier because salt "freezes out" from ice.
  • It is quite possible that in some areas closer to poles, salt would centralize in few areas, creating very salty lakes (which would not freeze even in winter, warming them up because of accumulation of glaciers in Arctic (2 miles thick above sea level like Antarctic), sea level in tropical areas would be much lower (some 100-300 meter less) because all that water would be solid in thick polar glaciers. Glaciers would extend much farther from poles, and climate would be colder (more sun energy would be reflected back to space).
  • ice reflects more sun energy than water, positive feedback). Effect in Arctic Ocean would be perma-ice all the way to the surface, even way above sea level in tropics, with occasional very salty lakes high in the perma-ice/glacier mountains. So perma-ice would be like solid land, supporting glaciers and salt lakes (with possibly addition of layers of dust brought by winds).
  • In snowball phase of Earth, oceans would freeze completely (bottom to top in winter) except near underwater volcanoes. As a result: Life would adapt to to survive occasional freezing, and be able to continue after thawing.
  • Re @Taemyr comment about lover albedo of ice. As @David Richerby correctly notes, if ice fell to bottom, if cold lasts long enough, body of water would freeze bottom to top. Then, even during summer it might not melt all the way to the bottom (ice being protected by cold water on top of it), and ice would accumulate until frozen solid bottom to top. Not good for complex life.
  • Effect on plant life on continents would be minimal. Seasons would be more pronounced, but plants can handle that. So coal should be possible - and also industrial revolution.
  • Such adaptation to survive many freeze/thaw cycles would be boost for slower-than-light travel.
  • So if advanced life and space-faring civilization could develop on such planet, they would have great advantage in space travel.
  • 2
    $\begingroup$ One winter when I was growing up there was a huge sleet storm followed by intense cold. All the roads were covered with inches of ice. When spring came, the ice was still in the streets with a sheen of melt water over it, and the city had to send equipment to all the streets to break the ice and pile it beside the roads because it would not melt by itself! $\endgroup$
    – Oldcat
    Jan 5, 2015 at 23:17
  • 3
    $\begingroup$ I don't think you would get a snowball earth effect if ice did not float. The runaway effect that drives drives snowball earth is that ice have a higher albedo than water meaning more heat is reflected. If ice does not float less earth would be covered by ice, and the albedo change would be less significant. $\endgroup$
    – Taemyr
    Jan 6, 2015 at 10:21
  • 1
    $\begingroup$ (1) Evolution has no goal, it only select best organism adapted to changing environment, so it cannot "evolve towards higher CO2". Evolution does not work that way. (2) Re: Frozen memories: maybe, maybe not. There also is life without freezing possible closer to volcanoes and in salt lakes. But yes, as I said in my answer, chance of more complex life evolving are lower (but not zero). $\endgroup$ Jan 6, 2015 at 14:57
  • 2
    $\begingroup$ @Taemyr In the real world, ice floats so lakes and seas freeze from the top down. The layer of ice insulates the liquid below so most of the water stays liquid and the amount of freezing is mostly governed by how cold it gets. If ice didn't float, freezing would happen from the bottom up. The amount of freezing would be governed by how long it stayed cold, since any drop in surface temperature below freezing point would cause more freezing. You'd probably find whole bodies of water freezing solid and staying frozen most of the year, increasing albedo a lot. $\endgroup$ Jan 6, 2015 at 15:15
  • 1
    $\begingroup$ @PeterMasiar Would all of the above apply to exoplanets with oceans of something other than water? For instance, would this describe an earth-analogous methane-ocean planet? $\endgroup$ Jun 28, 2017 at 14:21

I have a hard time answering this question because I'm stuck on how to make ice not float. Are we looking to rework the laws of physics themselves, or a work around within physics to justify it?

The easiest way to make this happen would be to assume a world where water was not the most common liquid. However, the question of "how would the world develop if some other liquid was the building block of life instead of water" would be far more important then this one.

Thus lets assume that were talking about water. The easiest way to keep ice from floating would be to decrease the density of water by contaminating it.

Going with this approach you would also have significantly lowered the freezing point of water. Thus there would be far less ice being formed to begin with. When the ice did begin to be produced it would sink down, allowing more ice to form. Presumably the temperature of water below the surface would be lower then above the surface, so the ice would continue to sink. As it sunk pressure on the ice would increase, which would further encourage it to freeze. The ice would not unfreeze easily at this point!

Thus, despite the freezing point of ice being much lower in these oceans, I imagine you would have huge frozen patches of ice in the ocean. Since once the ice sank the pressure and increasingly lower temperature would conspire to keep it sunk. I imagine that in fact most of the ocean would be frozen this way, only the very top of the ocean, where sunlight penetrates enough to keep temperature's above freezing, would be liquid. Luckily since the actual freezing point of water would presumably be much lower the ocean would never freeze entirely. Even during winter It would be too warm near the poles for water to freeze.


To add to this ice would be formed under the surface instead of at the surface of water. As of now in a perfectly clean lake that is frozen over will have a temperature of 4 degrees Celsius at the bottom. This is because 4°C is the temperature at which water is the most dense, after this point good old hydrogen bonds kick in and make colder water start to float back to the surface( quick tip, if your in chemistry class and asked why water does something funny assume it's hydrogen bonds, your ace every test). Thus the temperature of even huge bodies of water are pretty well regulated, no matter how deep you go your not getting colder then 4°C. Without this principle the water below the surface will be much MUCH colder since there is no source of heat and nothing to encourage the warm water above to sink lower.

This also means that water will 'churn' less, warm water won't naturally be pulled down further into the ocean. This probably means less oxygen in the depths, I imagine part of the oxygen that low is transferred from warm water on the surface sinking. I don't know how significant this effect is, probably minor compared to the effect plankton has. It's a moot point though, since water won't get nearly as deep before ice forms.

You would also have Ice poles, locations where the temperature of water is regularly lower then freezing where you have a solid 'pole' of ice from surface all the way to the ground. These poles would expand out radially until they reach a point where the average temperature is warm enough that surface water doesn't freeze. Then ice will slowly drop as you expand further out, from being on the surface to being 1 inch below to 3 inches below surface to 10, etc. etc. A slow steady drop as you expand radially out from the ice poles.

At warmer areas your have most of your ocean frozen. Those parts close enough to sunlight, and thus heat, will stay liquid water. I'm not sure how far you will have to go below the surface before you start hitting ice. I'm too lazy to do the math. However, I don't think it will be too deep. Lakes will probably be ice-free during the summer, but once you get deeper then your average lakes that ice will form due to lack of heat from sunlight. Your ocean life will be far less common.

Small lakes will freeze entirely during the winter, but perhaps not large ones. The water itself works like a huge heat sink, storing large amounts of heat. It takes time for that much heat to be expelled. And it takes a significant amount of heat to freeze water into ice to begin with. Thus large lakes may only 'mostly freeze' because the temperature warms up before the lake has had time to freeze over properly. To give an example of this concept, in reverse, imagine a pile of snow created by a snow plow. After it warms up above freezing snow will start to melt, but the snow in the snow pile could take a week or more to melt in full despite it being above freezing because the snow insulates itself. The same principle, in reverse, means it would take quite awhile for lakes to freeze. When the do freeze they would freeze from the bottom up, so large lakes would still stay liquid for most or all of winter.

However, any pond that is small enough to freeze over entirely will kill most animal life, though I stress most. Currently fish and other lake animals live underneath the ice where it's still liquid, the liquid ice on the top prevents the rest from freezing by insulating it so well. With the entire lake frozen it would seem that all creatures would die, but frankly that doesn't give evolution enough credit. Instead we would have fish that lay eggs that can survive freezing and which hatch after the lake thaws, fish that can live in shallow 'mostly frozen' lakes by hibernating, all kinds of ingenious evolutionary tactics to survive. Still, I would expect far less diversity in such ponds.

If you want a better analysis on the effects of this I would really encourage you to ask here: https://what-if.xkcd.com/ This is exactly the sort of question he answers, and he will do a far better job then I would :).

  • 7
    $\begingroup$ "Ice still floats in salt water, but not nearly as well." - does not compute. Salt water is denser than fresh water, and when ice forms, salt accumulates into droplets called brine, which are typically expelled back into the ocean, and what was left in brine pockets drains out over time, too, leaving air pockets behind. Ice floats better in salt water! $\endgroup$
    – Kreiri
    Jan 5, 2015 at 16:40
  • $\begingroup$ if ice didn't for hydrogen bonds then a lot of interesting chemistry probably wouldn't exist either $\endgroup$ Jan 5, 2015 at 16:51
  • $\begingroup$ @Kreiri Yes, you are totally write. I was thinking one thing and typed something completely different. I'm quite embarrassed by brain fart, removed it :) $\endgroup$
    – dsollen
    Jan 5, 2015 at 16:52
  • $\begingroup$ "I'm stuck on how to make ice not float" -- make the ice out of heavy water? :) $\endgroup$
    – Brian S
    Jan 5, 2015 at 22:23
  • $\begingroup$ Someone suggest Randall get a StackExchange account! $\endgroup$
    – Nefer007
    Jul 23, 2015 at 13:01

In our universe ice is less dense than water

  • Warm water is less dense than cold water and so will rise above it. The surface of the sea is usually warmer than the deep ocean. Warmth is exported upwards, and cold downwards. If this process continued unchecked, the sea would be uninhabitable.
  • However, this doesn't happen because ice is bizarrely less dense than water. This is because of its very unusual crystalline structure and so ice floats on water, even warm water.
  • Because ice is less dense than water, the deep oceans can never freeze, even below zero degrees because to freeze they would have to expand and the pressure prevents this.
  • Any ice that forms will float to the surface and be melted by the sun.

In an alternative universe where ice is more dense than water

Two things:

  • Ice would sink, and so would not be melted by the sun or rising warm water. Barring some heat source, the ocean would fill up with ice from the bottom up.
  • Ice would occupy less volume, so increasing pressure would raise the freezing point of water. The deep oceans would be permanently frozen, just as the core of Jupiter is frozen, not because it's cold, but because the pressure is so intense.

Depending on other souces of heat (geothermal, solar, etc), it's possible the entire sea might freeze from the bottom up.

Alternaltely you might end up with a thin layer of warm liquid ocean a few hundred meters thick atop a giant ice sheet with mountains of ice rising at the poles.

Neither option would be good for us humans.

  • $\begingroup$ Note that the maximum density of water is actually at 4C, which makes the picture a little more complicated, though I think your ultimate conclusions remain sound. $\endgroup$ Jan 6, 2015 at 15:19

Currently here in the north we are aware of lakes turning over, this happens because water gets denser as it cools to a certain point (according to google 4C or 39.2F) then it starts getting less dense. So when water cools to 39 degrees it sinks and the warmer (or cooler) water rises. Other than having currents in the water and helping move oxygen down and nutrients back up I don't remember what else it does. But the Oxygen I think helps the fish survive the winter with the Ice on top. The Ice would sink evening out the temperature of the water all the way to the bottom of the lake. Lakes would freeze from the shore out to the center from the bottom up. Likely making it easier to freeze a lake out, since right now the as the Ice forms is slows the process of freezing more water, like putting a cover on a pot of water, to slow the escape of heat. Small lakes in the north might freeze through and take most of the short summer to thaw back out.

However, if ice sank I would have to question the speed at which it would melt? There are two things to think of here. 1 the average ground temperature about 5-6 ft. down is I think 52 degrees F. so the ice that settles will likely slowly melt.

You also have pressure, currently normal ice melts under pressure, there are several different kinds of ice that are created under pressure, I would expect large parts of the ocean bottom would be covered with these different types of ice to some unknown depth. Currents might dig them deeper in places.

This I can see would have a large impact on our eco-systems. If it was a sudden change and not something we've always had it would be devastating to the entire planet. The north pole ice cap would sink and be open for trade. Everything under the Ice cap would likely die, except maybe for some micro organisms.

The sinking of the Ice cap would cause a huge tidal wave out then back in. but being denser than the water it would collapse in is size and likely end up even LOWERing the ocean levels (by just a little bit). Ocean currents would be completely messed up, and all the doom and gloom that is predicted by the melting of ice caps and stopping the currents will come about. A very large part of the planets Oxygen is produced by plankton in the ocean and this could be severely disrupted. Seriously endangering the planet with not enough CO2 -> O2 production. Quite possibly leading to global warming and most of the issues related to that.

My conclusion, it would not be good at all.

  • 1
    $\begingroup$ Wouldn't floating ice would have only a small effect on sea level as it sunk since it's already displaced it's weight in water? $\endgroup$
    – Tim B
    Jan 5, 2015 at 17:00
  • $\begingroup$ @TimB yes, melting the north ice cap, would be a zero sum game. Making the north ice cap denser to sink, should reduce the total volume and actually lower the ocean levels just a little bit, a large part would depend on how much denser the ice is. Antarctica is a completely different issue. $\endgroup$
    – bowlturner
    Jan 5, 2015 at 17:04
  • 1
    $\begingroup$ @TimB The problem is the actual sinking of the icecap. take a big stone and hold it level with a pond and then let it go. see what happens. $\endgroup$
    – bowlturner
    Jan 5, 2015 at 17:05
  • $\begingroup$ true, but it would depend on just how much heavier it became - a more gentle sinking is entirely possible. $\endgroup$
    – Tim B
    Jan 5, 2015 at 20:06
  • $\begingroup$ it would not freeze as a sheet then sink it would have an effect more like snow but underwater. $\endgroup$
    – John
    Feb 18, 2021 at 1:30

Ice is less dense than water because its crystalline lattice is kept together by hydrogen bonds.

At about 4°C hydrogen bonds begin to form, even if they're not stable before freezing and thus water molecules start to form bigger (and unstable) aggregates that will later give rise to ice crystals.

Hydrogen bonds are so basic in (almost) all chemical processes within cells that nothing would work without them: DNA helices would not pair, protein chains would not bend the way they do, "respiratory chains" in mitochondria wouldn't work, etc. In short: life as we know it wouldn't be possible.

Note: I know You asked a different question, but you cannot separate cause from effect.

Note2: In case it was not clear, the answer to the question is: "a world where ice is heavier than water would not have any life as we know it (among other things)." I added the answer because I felt this is a relevant difference that was not included in other answers.

  • 1
    $\begingroup$ While interesting and a good point, as you say this doesn't really answer the question. $\endgroup$
    – Tim B
    Jun 27, 2017 at 15:53
  • 1
    $\begingroup$ One of the questions: "How would ... life be affected?" This answer: life wouldn't work at all without the property that causes ice to be less dense than water. So unless you change from water ice to something else, it wouldn't work. $\endgroup$
    – Brythan
    Jun 27, 2017 at 18:48
  • $\begingroup$ @TimB: I edited the answer to make it a bit more explicit. $\endgroup$
    – ZioByte
    Jun 27, 2017 at 18:52
  • $\begingroup$ @ZioByte You are assuming though that the change would be disruptive, there could be other mechanisms causing the density inversion. Alternatively if life had evolved on a world with water that behaved like this then life would be adapted to it. $\endgroup$
    – Tim B
    Jun 27, 2017 at 20:16
  • $\begingroup$ @TimB: Sorry, I do disagree. "if life had evolved on a world with water that behaved like this then"... it would be like nothing we can imagine, not a simple "adaptation". This is very fundamental. Try having a computer with no "bits"; You can actually do that (analog computers) but it's literally another world. Even if You can do the same computation with both they actually have (almost) nothing in common. $\endgroup$
    – ZioByte
    Jun 27, 2017 at 22:17

First of all, water does contract as it freezes at very high pressures, well above the critical point. If we say that water behaves in this way at a standard atmospheric pressure, then bodies of water would freeze from the bottom up.

To determine the behavior of oceans, we would also have to take into account the effects of high pressure. The melting temperature of all substances with a denser solid than liquid phase increases with increasing pressure; think of the substance as being squeezed into its denser state. Just how much the melting temperature would increase would depend on how much denser the ice is than the water, but assuming that at 1 atmosphere it has a melting temperature of 0C and that temperature increases with pressure at the same rate that it decreases with pressure in our universe, you could have room temperature ice about 2-3 kilometers down; the depth of the liquid portion of the ocean would depend on temperature and salinity, but even for very warm water would probably not extend below 5-6 kilometers (which is a fairly typical depth for an ocean).

In cold areas it could of course freeze all the way to the top. There would probably be a layer at the ice-water interface that melts and refreezes seasonally or with changing currents; below that the ice would essentially behave as a mineral with a hexagonal crystal structure, thermal conductivity and hardness well within the typical range for minerals, and the peculiar property that it melts if you bring it to the surface. Since you would effectively take away convection and radiation as heat transport mechanisms, temperature gradients within seafloor ice would be conduction dominated and it would increase in temperature with depth just like rock does; despite no sunlight penetrating to the lower ice layers, it would be quite warm, just as rock at depth is warm or even hot.

Underground reservoirs of water would behave in a similar way, acting as mineral veins instead of a fluid. Mining through ice would present interesting problems, as it would revert to a liquid state as soon as you relieved pressure by digging down to it (not all at once probably, because of latent heat, but that would also make it very cold).

In terms of overall climate, a world like this would probably have greater seasonal temperature swings because of shallower oceans, but not by a huge amount. I expect the geology would be quite different, with water acting as a structural element below a certain depth and anything dependent on marine biology, such as chalk and oil, being deposited in ice rather than seafloor mud.

You'd get all kinds of other quirky effects; one of the reasons ice is so slippery is because a thin layer melts when you step on it, but of course this wouldn't happen if it freezes harder with pressure, so you might get better traction on ice. Of course, if you're going to change the nature of water anyway, you may as well mess with any of the properties I cited and get completely different results.

  • 2
    $\begingroup$ At present, this is somewhat of a wall of text which makes our eyes bleed. Overall a detailed answer to what would now be considered an off-topic question. Please edit for readability (paragraphs are important). Enjoy our tour and when you have a spare bit of time, read-up in the help center about how we work. Welcome to worldbuilding Thomas. $\endgroup$ Feb 18, 2021 at 1:25

You must log in to answer this question.

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