Water on Mars: Where the Ice Actually Is (and Who Gets It)

Here is the most underrated fact about Mars: it is not a dry planet. It's a frozen planet we keep calling a desert because the deserts are the parts we landed on.

There is enough water ice on Mars to cover the entire globe in a layer somewhere between 20 and 30 metres deep, if you melted all of it and let it settle. That comes from decades of orbital radar, neutron spectrometry, and the occasional lucky landslide that exposes a wall of clean ice like a cut cake. The problem was never quantity. The problem is that almost all of it is in the wrong place, at the wrong latitude, and everybody wants the same few good spots.

So let's go find it. This is a dispatch about Martian real estate, and on Mars, real estate is water.

The polar caps: enormous, obvious, useless

Start with the answer everyone knows. Mars has two polar caps, and the northern one is mostly water ice — a stack roughly 1,000 kilometres across and up to 2 kilometres thick, laid down in visible annual bands like tree rings. The southern cap is a stranger beast: a permanent carbon-dioxide lid on top of a much larger water-ice mass, plus whatever radar keeps arguing about underneath it.

Together the caps hold most of Mars's known water. And no serious colony plan puts the first city there. Why? Sunlight. A polar base gets months of darkness, brutal seasonal swings, and terrible orbital access — every kilogram launched from high latitude fights the planet's rotation instead of borrowing it. You'd be founding a fragile settlement in the Martian equivalent of Antarctica's interior, except Antarctica has breathable air and a rescue plane.

The caps are the strategic reserve, not the founding site. Which means the actual question — the one that decides where humanity's second civilization gets built — is: how far from the equator do you have to go before you hit ice?

The mid-latitude glaciers: the real prize

The great discovery of the last twenty years is that Mars has buried glaciers — not polar ice but mid-latitude ice, in the 30° to 55° bands of both hemispheres, under a protective blanket of dust and rubble a few metres thick.

Orbital radar found them by listening to the echo: dust and rock return a messy signal, clean ice a suspiciously honest one. Some of these lobate features are hundreds of metres thick, and some are so pure they're effectively glaciers wearing a dust coat — preserved because Mars's wobbling axial tilt once made those latitudes cold enough to deposit ice, then buried it before the sun took it back.

That dust blanket is the whole story. It means the ice survives at latitudes with tolerable sunlight and decent launch geometry. It means you can build a base at 40° north and dig for water instead of trekking to the pole for it.

And in a few places — steep scarps where the ground has collapsed — that ice is exposed in cliff faces, sometimes starting a metre or two down. Those cliffs are the most valuable land in the solar system that nobody owns yet.

The equatorial disappointment

The equator is where you want to land: warmest, sunniest, best launch physics. It is also where the ice mostly isn't — Martian conditions sublimate exposed ice straight from solid to vapour, skipping the liquid phase entirely.

Not a total loss. Equatorial Mars has hydrated minerals — clays and sulfates that bind water into their crystal structure, laid down by ancient wet environments. Bake them and they give it up. The catch is efficiency: you might process a tonne of regolith for a few percent of its mass in water. That's a mining operation, not a tap.

Follow the water — the organising principle NASA adopted for Mars exploration, on the logic that water is where the habitability question, the biology question and the fuel question all converge.— paraphrasing NASA's Mars Exploration Program strategy

What the water is actually for

People assume water on Mars is about drinking. Drinking is the smallest use by volume and the easiest to recycle — closed-loop life support on the ISS already reclaims most onboard water, and a Mars base would run tighter. The real consumers, in rough order of appetite:

Rocket fuel. Electrolyse water into hydrogen and oxygen, feed the hydrogen and Martian atmospheric CO₂ into a Sabatier reactor, and you get methane and more water back. Methane plus oxygen is a return ticket, and a single crewed return vehicle needs propellant measured in tens of tonnes — a small pond.

Agriculture. Every calorie grown on Mars is water circulating through soil, plants, and air. It's recoverable, but the standing volume in the system is large and scales directly with population.

Radiation shielding. Water absorbs the energetic particles that Mars's thin atmosphere fails to stop. A few metres of it around a habitat does what a few metres of regolith does, and unlike regolith you can drink it in an emergency. Every wall in our tour of a Martian metropolis in 2100 is doing two jobs.

Industry. Concrete, chemistry, cooling, cleaning. Civilisation is a wet process.

Why the first water claim is the first political crisis

Here's where we get knowingly unrealistic, which is the house style.

The good deposits — thick, pure, shallow, at a survivable latitude, near flat landing ground — are a short list. Maybe a few dozen sites on the entire planet clear every bar at once. The Outer Space Treaty says no nation may claim sovereignty over celestial territory. It says considerably less about what happens when a company parks a drilling rig on the only clean ice cliff within 400 kilometres and calls it an operational safety zone.

My forecast: the first genuine Mars conflict is not about independence or ideology. It's about a water table. Two outposts, one buried glacier, a supply window 26 months wide, and a legal framework written in 1967 by people imagining flags rather than pumps. The independence arguments come later — we sketched that escalation in the century-by-century forecast — but water comes first, because water is what everything else is made of.

The honest uncertainty

Everything above is built on remote sensing: radar echoes, neutron counts, spectral signatures, a handful of shallow scoops. What we do not have is a single deep drill core from a mid-latitude glacier.

So the numbers could move. That "hundreds of metres thick" ice sheet might be 60% rock. That pristine scarp might be brine-saturated and corrosive to every pump we bring. The first colony's water yield will surprise somebody, and the surprise will either be a boring engineering delay or the most important discovery in our history — because wherever there's persistent liquid water, the question stops being about plumbing. For why we keep going anyway, that's a whole dispatch of its own.

Mars is a planet with oceans. They're just filed inconveniently.

Frequently asked questions about water on Mars

Is there liquid water on Mars right now?

Almost certainly not stable liquid water at the surface — low pressure and temperature make exposed water sublimate directly to vapour. Radar hints at briny liquid deep beneath the south polar cap, but clays and frozen brines fit the same data. Salty transient films in warm soil remain plausible and unproven.

How much water is on Mars in total?

The known ice — polar caps, buried mid-latitude glaciers and ground ice — works out to a global-equivalent layer of roughly 20 to 30 metres if spread evenly over the planet. The difficulty is accessibility, not quantity.

Where would a Mars colony actually get its water?

Most serious architectures target shallow buried ice in the 30°–50° latitude bands, where dust-covered glaciers sit within a few metres of the surface and sunlight is still workable. You'd mine ice directly rather than bake hydrated minerals — the energy cost per litre is dramatically lower.