Consider a scenario where all freshwater on Earth suddenly disappeared...

Would the planet remain habitable?

Definition of water removed:

  • Everything in this chart under the category 'freshwater', except living things.
  • 14
    $\begingroup$ Do you also intend glaciers, non-glacial snowpack, and water stored in bottles and tanks to also disappear? $\endgroup$
    – user535733
    Jan 11, 2018 at 0:44
  • 9
    $\begingroup$ So, a bottle of coke is dry, but watermelon is juicy? $\endgroup$
    – Alexander
    Jan 11, 2018 at 1:01
  • 4
    $\begingroup$ Is salt water freshwater in this scenario? $\endgroup$
    – Brythan
    Jan 11, 2018 at 3:09
  • 11
    $\begingroup$ Define “fresh water”. Is a fifty litre water tank with a teaspoon of salt (5 ml) in it affected? Ten teaspoons? A hundred teaspoons $\endgroup$
    – Mike Scott
    Jan 11, 2018 at 9:55
  • 9
    $\begingroup$ Does this include water vapor? Pretty sure it is always raining atleast somewhere. $\endgroup$ Jan 11, 2018 at 14:03

9 Answers 9


Depending on the temperature you are exposed to, you can go 100 hours without drinking at an average temperature outdoors. If it’s cooler, you can go a little longer. If you are exposed to direct sunlight, it’s less. (source)

Which is a fancy way of saying almost everything on land is dead, and that's just the beginning

Even life in deserts depends on some fresh water. Aquifers would quickly be flooded with sea water that would take years, maybe centuries to purge. The oceans account for 96.5% of the Earth's water, but it's useless for anything but fish. That means survival is 100% dependent on rainwater and the restoration of the water cycle.

With no humidity in the air it will take time to re-hydrate the atmosphere. It might (maybe!) take days to get rain to fall on islands, but it'll likely take weeks to get rain on any mainland but the coasts. The rain would be soaked up by the ground instantly for weeks beyond that before rivers and streams formed that outlast the rainstorms.

The earth would instantly have new and expanded deserts. My guess is that the entire U.S. great basin would become a desert (as an example) with little to no hope of recovery due to the loss of fresh water.

Icebergs are primarily fresh water, so you just lost the polar caps, which means a significant source of world cooling just disappeared. Worse, the freezing process (not instantaneous by any means, but still significant) is starting to draw fresh water from the oceans. I don't know how much this would increase salinity — probably not a lot — but not a lot is needed to start killing the fish.

The reality is that people (much less animals) couldn't capture enough rainwater fast enough for anywhere near the existing population to survive. Everyone not on a coast is dead, guaranteed. Densely populated coastal areas are dead. Frankly, I don't believe rain would fall in significant amounts within the proposed 100 hour limit. That leaves evaporating and condensing sea water, a tedious process that couldn't be put together on behalf of a lot of people fast enough. Individuals, maybe families would survive... But city dwellers would be in a world of hurt.

  • People would, if they can, evaporate sea water and recondense it for drinking water. I can see this for families. Maybe small communities. I can't see this happening for cities fast enough to save the majority of the population.

  • People would, briefly, resort to vampirism (I am not kidding). They'd drink the blood of anything they could get their hands on for as long as the supply lasts or is replaced by some other solution.

  • Most of the complex land-based life forms are dead. Plants would eventually recover.

  • You might lose a fair amount of seafood thanks to the salinity change. That means people just became vegetarians.

I'll be honest with you, the more I think about it, the more cataclysmic this becomes. There would be survivors, but life would reek.


A conversation in the comments brought up an interesting point. Would the power grid stay running long enough to power desalination on a large enough scale to help survivors?

Yes and no. For example, hydroelectric only represents about 6% of the total power in the U.S. However, the remaining power grid must make up for that loss - and it will be forced to make up for it because unless everything is shut down in a microsecond, the grid will forcefully yank it from elsewhere to fill demand (that's how electricity works, the grid is actually designed to accomodate this).

6% doesn't sound like much, but it's an enormous variance across the entire U.S. grid. Without a chart, it's a bit difficult to explain, but imagine a happy electrical grid as a straight line. Suddenly those chunks powered by Hydroelectric drop out... a step signal (square wave) forms in that straight line between the line and zero ... and starts moving to fill in the areas forced to zero. When it's filled, it doesn't stop. It slaps against the "end" point and reflects backward. Now you have a square wave above the line moving out. Back and forth and back and forth until the disruption is finally aborbed.

In the sections around the grid nearest those powered by hydroelectric, this is catastrophic, tripping breakers to protect the "backside" of the grid, but those breakers aren't placed to protect the grid in the event all the water disappears, so some of those breakers will shut down the mains from other power generators, like coal. So the area of effect is bigger than just 6%. How much bigger? Without analyzing where all the breaker locations are, that's a complete guess.

However, because electricity "moves" faster than breakers, even areas well beyond the hydroelectric parts of the grid are effected. That reflective "bounce" I just described can burn things out. So you'd see appliances crashing all across the nation. (The more delicate the more likely the crash. Protection by a power backup is VERY important as it would protect you from this.)

Finally, who uses hydroE? The northeast costal towns would be dead, and there's a better-than-average chance all coastal towns would be affected. Check out this map of where the hydroelectric grids are.

But the real problem is steam generation. Even nuclear and coal plants are basically just big steam turbines. No water, no steam, no electricity at all... within seconds.

Frankly, this would be an uber-cool addition to your story.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ Jan 14, 2018 at 3:58

In addition to the biological consequences discussed above...

From what I can tell, at any given time, there are very roughly 30,000 cubic kilometres of sea ice in the Arctic and Antarctic combined. If that ice suddenly vanishes, it creates a 'hole' in the ocean of 30,000 cubic km (or if you want to be fussy, 30,000 km divided by 1.04 or so). Salt water will rush in to fill that gap. I would expect MASSIVE tsunamis to result.

How massive? Well, about eight thousand years ago, a large chunk of oceanic collapsed in what's known as the Storegga slides. "Some models of the Storegga slides estimate that tsunami waves exceeded 20–25 metres (65–80 feet) in height along the coast of the Shetland Islands, 10–12 metres (33–39 feet) along the Norwegian coast, and 5 metres (16 feet) along the coast of eastern Scotland."

The submarine landslide that caused Storegga involved a total of about 3500 km^3 of collapsing sediment. The volume of ice you're magicking away would be about ten times as much. While the two scenarios have other differences that might affect the severity, I think it's safe to say that a large chunk of the world's coastline would experience massive tsunamis.

Meanwhile, you're also losing the ice sheet on Antarctica, Iceland, and various other places. It's believed that a thick cover of ice has a significant role in suppressing volcanic activity in Iceland; without it, you can expect an increase in eruptions. Meanwhile in Antarctica you're suddenly exposing a continent's worth of glacier-ground dust to blow out over the world.

edit: just thought of one more thing. Volume of ice cap is ~ 3E7 km^3 = 3E16 m^3.

If that vanishes suddenly, and air rushes in to fill it... the energy involved is about 3E16 m^3 * 101.3 kPa = 3E21 joules. Which is about the same as a million megatons. For various reasons, won't be exactly like a million MTs of nuclear explosions, but it's still colossal. Expect huge amounts of dust sucked into the air; you probably have a super-nuclear-winter type of effect to look forward to.

I don't like anybody's chances.

  • 1
    $\begingroup$ Wow -- hadn't thought of this. Starting to look like the only realistic chance of survival in this scenario is to be in a lightly populated area, within walking distance of an inland or otherwise sheltered sea, but not too close to it to start with. There can't be many such places in the world. $\endgroup$
    – Jules
    Jan 11, 2018 at 7:06
  • 1
    $\begingroup$ This would be the real problem. If you want to see the bright side of life ... tsunamis transport water on land to desalinate naturally and volcanic activity boils seawater aswell ... $\endgroup$
    – Thaylon
    Jan 11, 2018 at 13:55
  • 7
    $\begingroup$ Antarctica, losing that much ice mass, would pop upwards. And I mean pop. You don't expect continents to "pop", but removing 28 million km^3 of ice at once is not an expected event. I'd guess that the crust will crack and the entire continent will experience a volcanic eruption. $\endgroup$
    – Yakk
    Jan 11, 2018 at 16:31
  • 1
    $\begingroup$ @Yakk yeah, I'd certainly expect some exciting seismic consequences, but I don't have a good feel for how quick that would be. (If the ice vanishes REALLY fast, we also have a big atmospheric shockwave to deal with...) If we knew something like this was on the way, we'd probably want to melt the ice caps as fast as we could! $\endgroup$ Jan 11, 2018 at 21:45

The most obvious answer is that you wouldn't have any impact on human society because humans themselves would cease to exist. If you take the water out of living things as well as out of the water cycle, then humans would all be dead as well.

The human body (like most organic creatures) is around 70% water (depending on which science journal or estimate you read) and we simply can't survive without that water in our bodies. We'd be dead. Instantly.

Ignore this part; I misread the first paragraph of the OP and thought that the water in living things was also gone. That of course creates a much more catastrophic scenario.

For the sake of argument, let's assume that all water is gone except that in human bodies.

We're still dead, but it would take a couple of days.

The human body needs around 2L of water per day to survive. within 3 days or so, we'd all be dead of dehydration.

There are some exceptions in this case; desalination isn't that hard to achieve if you know what you're doing. Some humans could survive by rigging up freshwater stills at the shores of oceans, but by now society has descended into the chaos demanded by Maslow's hierarchy of needs and defending your supply of water would be the real trick. Some other humans may try to survive by drinking the blood of other humans, which isn't going to solve the problem but is still going to be tried. Still others will eat some extra salt to retain what water is in their bodies already, but this is also an extremely short term measure.

If all this water just disappeared from the water cycle, then ocean levels would drop, because there's no freshwater rivers running into the ocean to replenish what gets evaporated and turned into clouds above us. Remember, that the water in clouds is fresh water (effectively distilled by the evaporation process) meaning that in your event, all clouds have gone as well. They would form again in time, but you're probably talking a few days.

This is good for the desalinators, as there is plenty of sun to create their own distilled water. But, it's going to make things hot across all of the earth, meaning that those without water will last for less time than they would if it was cooler (less sweat).

The bad news is that rivers, lakes, even streams don't form again for a long time. They are based on massive amounts of water falling in key mountainous regions; in many cases as snow that melts at a steady and sustained rate. In short, depending on the time of the year this happens and even assuming no long term disruption to existing weather patterns (there would be) you'd be talking up to 2 years before the water cycle could start to look somewhat similar to what it did.

By then, everything would be already dead.

All life on Earth requires water in some form. It's possible that some hardy plant seeds could go dormant during that time and wait for moisture before germinating, but the bacteria in the soil they rely on to grow would already be dead as well. So, it's unlikely that they would grow to a point where they could reproduce.

In short, this would be a catastrophic event for all life on earth and I strongly recommend that you don't try doing it.

  • 2
    $\begingroup$ If the catastrophe includes instant vanishing of sea ice, those coastal communities may well have to survive a major tsunami before they can think about setting up their stills... $\endgroup$ Jan 11, 2018 at 4:03
  • 1
    $\begingroup$ @GeoffreyBrent; that's a good point. Ultimately it means that animal life would probably die out sooner and those in the coastal regions are ultimately lost from the Tsunami, leaving people slightly more inland with the slim hope of migrating inland to set up stills in time. It's still pretty bleak. $\endgroup$
    – Tim B II
    Jan 11, 2018 at 6:01
  • $\begingroup$ Hmmm. I'll crack gasoline for water. $\endgroup$
    – Joshua
    Jan 11, 2018 at 17:27

An aspect not mentioned in the other answers is the geological calamity. Aside from most organisms dying of thirst of the next few days, a lot of the planet would probably shift or collapse. The structural stability of a lot of landmasses depends on the subsurface aquifers being full. With those suddenly being empty, expect a lot of sinkholes and general shifting of the surface as internal cavities collapse. Beyond that, there is (probably) a lot of water trapped in the earth's mantle in the form of hydrous silicates. Depending on the specifics of your scenario, this might seriously alter the convection currents in the mantle, possibly triggering earthquakes (crust plates ride on those currents) or volcanoes.


All water

Total amount of water in the world: 1,386,000,000 cubic kilometers
Total amount of water in the world: 1,386,000,000,000,000,000,000 liters
Total amount of water in the average human: 42 liters
Number of humans: 7.6 billion
Total amount of water in humans: 319,200,000,000 liters or .3192 cubic kilometers

So we'd have a huge decline in the total amount of water. As a practical matter, there wouldn't be enough water left to grow crops. The water in us would boil off as sweat and could not be replaced. The crops would dry out and die. Only people in closed loop recycled situations could survive.

Just freshwater

Now, if you literally meant just freshwater, then we're better off. Most of the world's water is salt water. So we only lose about 4% of the world's water.

For the first week or so, the air would be very dry. The days should be hotter and the nights cooler, as there are fewer clouds. But after a week or two, there should be enough evaporation to replenish the atmospheric water.

Total amount of atmospheric water: 12,900 cubic kilometers
Total amount of water that enters the atmosphere daily: 1170 cubic kilometers

Water should evaporate faster in the beginning, as the air is dryer. But there are fewer places with water, so by the end, it should be slower than previously. This might leave the air somewhat drier but not a lot.

After the first couple weeks, we should have about as much precipitation as evaporation. But only some of it will fall on land. Let's figure about 30% (the world is about 70% ocean). This may be generous, as most of the moisture will be coming from the ocean. It's not clear how far inland it will take it. Anyway, call it 330 cubic kilometers a day.

Freshwater: 10,633,450 cubic kilometers

That's about 30,000 days of precipitation to restore the freshwater. However, we'll also be losing water to evaporation at the same time. And the difficulty getting water inland. So we're talking a century or two. Things might not be back to what we consider normal now, but they will be mostly in balance. There would be more inland deserts. These might take millions of years to disappear, as they soak up any precipitation into groundwater. Of course, that means that they don't generate precipitation from their own evaporation.

The source of the numbers is the water in the world link at the top of the post.

Billions of people would die in that first week, with no source of water except food. More would die in the next couple months, as disease would spread rampant due to the lack of sanitary hand washing. Billions would die over the next year due to crop death during the week and an inability to replace the crops that died. This could easily lead to only millions of people worldwide.

Don't forget things like nuclear power plants. They rely on water for cooling. They also may require water for shutdown. Some flood the nuclear chamber or immerse the rods to stop and cool the nuclear reaction. Plants may meltdown or require catastrophic measures that ruin the plant. No restart.

Cars also need water to run, but perhaps the antifreeze saves them.

Places with desalination plants may be in the best shape. At least they can provide drinking water almost immediately.

  • $\begingroup$ Not just nuclear reactors, but volcanoes like Yellowstone require water to keep them cool. You would have nuclear meltdowns, and probably violent volcanic eruptions $\endgroup$
    – BlackThorn
    Jan 11, 2018 at 19:12
  • $\begingroup$ Desalination plants will be useless--they require power to run. The only generators beyond the home level that still work will be solar/wind--the grid crashes, you can't get the power to the desalination plants. $\endgroup$ Jan 13, 2018 at 5:10

As most have already pointed out this is catastrophic.

But some extra catastrophes not yet listed: Many dams need the water to remain stable. Without the water pressure they may structurally fail, not so bad though as they have nothing to leak, just will not hold water as it refills.

Many dams provide significant hydroelectric power. Lights out to those parts of the world.

Further on power, as Brythan pointed out all power plants need coolant,though many are coastal so may again do surprisingly well if saline water can be used in any regard, as well as some non water coolant solutions, but most internal are about to go bang.

Also high speed pumps need to pump mass else they will burn out very fast, if you just removed all the water instantly from a turbine it will probably burn out, if not undergo a rapid unscheduled disassembly.

This likely means all water grids just broke. Also the pipes may well cave in due to loss of pressure too. On cave ins... the emptying of ground aquifers will cause massive sinkholes over many limestone regions.

Some good news not covered: Some parts of the world rely on significant desalination plants for tap water. I guess significant parts of the Middle East may get off surprisingly well in the short term, some US states rely heavily on desalination and even Themes valley can supply something like 20% of demand which would easily cover London for rationed drinking water only.

Not so good news. Unfortunately the almost total global crop failure due to no rain and ground water means any significant population center will still starve even if water can be supplied, though then again there may be enough grain stored to cover those few surviving population centers for a couple years until reasonable crop recovery, though you are onto regular post-apocalyptic issues like the breakdown of civilization.

  • $\begingroup$ Ooooh. Rep for pointing out the hydroelectric problem! $\endgroup$
    – JBH
    Jan 11, 2018 at 18:13
  • $\begingroup$ Yeah +1 for the catastrophic structural failure of dams. $\endgroup$
    – Austin A
    Jan 11, 2018 at 18:29

As other answers have already pointed out, most of the world's water is salty, so the water cycle will generally carry on.

There's an interesting implicit question here. How "fresh" does the water need to be to escape the Great Water Vanishing? Many rivers such as the Ganges are profoundly polluted - do they stay around or vanish? Muddy soil is even more questionable, since you have typically a colloidal solution with more soil than water.

The most obvious outcome is that all freshwater aquatic species will immediately go extinct. Aquatic species with some amphibious capabilities may survive, but anything purely aquatic is pretty much dead. If water in soil survives, I'd say that's pretty much the end of it ecologically. Other answers have covered the human-scale problems, but these do not pose an existential risk to the current world as we know it.

If water in soil is taken as well though, you have a whole lot more problems. Fertile soil requires bacteria, microorganisms and generally an ecosystem, and it is by no means assured that they will survive dessication and rehydration. There will be substantial problems getting plants to grow properly in future.

More seriously, many climate systems rely on water to preserve them. Whilst water is present, they are in a stable equilibrium state - but take the water away and they will go to a different stable equilibrium state. Forested areas, and particularly rainforest, are very prone to this. Take away the cooling effect of evaporating water, and the water cycle in that area simply will not return to the same state. This has been proven in many areas, from sub-Saharan Africa to South America. The Earth generally will survive, but it is likely all species in those kinds of environment will go through an extinction event. Those that can adapt (or at least wait it out) will survive; but many will not.

  • $\begingroup$ The specific freshness has been provided in an edit to the question. $\endgroup$
    – Malady
    Jan 12, 2018 at 3:36

I am assuming the salt water would remain otherwise we would all die within a week or so. No story left to tell.

The extent of the damage would truly depend on what was classified as fresh water and/or water that would be taken.

Even so it would be terrible.

Many places on earth rest on large pockets of water, and with the water gone the ground above would start to crumble. Eventually massive, potentially city sinking sink holes would form.

The missing water weight would cause the ground to start rising up, just like after the glacier melted. Eventually this would cause earthquakes.

Some scientist say there are large pools of water at undetectable depths within the earth, and if they were fresh water and disappeared we would be in for even greater earth quakes as they collapsed in on themselves.

If you rip away all the ice bergs because they are fresh water, large parts of the land start to rise all at once because the mass of the ice is actually compressing the ground below it. The hardest hit beside the poles would be Canada and Russia and the ground rising could and probably would lead to earthquakes.


From Wiki:

Fresh water is naturally occurring water on Earth's surface in ice sheets, ice caps, glaciers, icebergs, bogs, ponds, lakes, rivers and streams, and underground as groundwater in aquifers and underground streams. Fresh water is generally characterized by having low concentrations of dissolved salts and other total dissolved solids. The term specifically excludes seawater and brackish water although it does include mineral-rich waters such as chalybeate springs.

Thus, some water would remain, from wet soil and air humidity to fruits, beer and even bottled water.

  • The eventual re-stabilization (?) of The Water Cycle.

The water cycle is actually not really interrupted. Atmospheric water (residence time 9 days) and soil moisture (residence time 1 - 2 months) are by definition not affected. Rivers would need 6 month, lakes and groundwater 100 years and deep groundwater 10,000 years to recover. Most plants and thus animals could however survive (obviously not fresh water fish ...).

  • Survival strategies that individuals/groups would apply.

Your best chance at survival is not being born in a region where water was scarce to begin with, or scarce enough that you already depended on desalination. Otherwise, simply consume existing supplies (food, plants, animals, other humans) and desalinate seawater.

  • Large-scale changes for the Earth.

Obviously, there will be interesting times. Anything depending on rivers has to adapt (or go extinct ..). Some plant and animal species may go extinct before nature has recovered, but there should be enough plant life thats mostly dependent on rain to support animals and humans.


Following the updated broader definition of fresh water, specifically including soil moisture and atmosphere, we're dead. (Even though it still doesnt include bottled water or other beverages.)

Disappearance of soil moisture kills off land based plant life quickly, essentially changing all soil to desert, which most plants cant survive. Lack of atmospheric water for several days assures that soil fauna is destroyed before it can regenerate. Even desert flora depends on atmospheric water, which is gone.

Global wild fires, decomposing plants and sand storms will pretty much make the air unbreathable.

A small percentage of humans could theoretically survive on desalination and seafood, but at least the enormous amount of unburied dead people within the first few weeks would help some species to survive a bit longer and others like flys to prosper.

Would the planet remain habitable?

No. The atmosphere would be overloaded with decaying bio mass and ash, lacking oxygen. The land would be a global desert. I'd assume, given enough time, this would impact sea life aswell, making human survival for the long term entirly impossible.

Would the planet become habitable again?

Probably. Some bio mass survives, and evolution has done its job once before with less options.

  • $\begingroup$ Are you suggesting that airborne water vapor, and better yet bottled water, are not affected because they are not "ice sheets, ice caps, glaciers, icebergs, bogs, ponds, lakes, rivers and streams [...] aquifers and underground streams"? Or are there other reasons too? While generous, and the people in the scenario would appreciate it if your definition were used, still... it seems like a stretch, even given your quoted definition, to suggest vapor is not affected, and like outright cheating to say bottled water is not affected. $\endgroup$
    – Loduwijk
    Jan 15, 2018 at 21:41
  • $\begingroup$ Yes, based on the wiki definition of "fresh water". Now, that the author has clarified ... see edit. $\endgroup$
    – Thaylon
    Jan 17, 2018 at 12:57

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