On Mars, you can stand in a place that looks unmistakably like a dry riverbed on Earth — branching channels carved into the rock, the kind of pattern water leaves when it flows downhill for a long time. There are ancient lakebeds, too, with sediments layered the way they settle at the bottom of a standing body of water. And there are minerals that only form when rock sits in liquid water for ages — clays, sulfates, carbonates. The catch is that none of this water exists today. Mars is bone dry on its surface, and so cold that any puddle would freeze or boil away within minutes.
That contradiction — a desert covered in the fingerprints of rivers — is the riddle this whole chapter turns on. Because if you want to know whether you could make Mars habitable, the single most useful thing you can study is the fact that it already was, and then wasn't. Nature ran the experiment for us, in reverse. It took a wetter, warmer, thicker-aired Mars and slowly turned it into the planet we see now. Understanding how that happened tells you exactly what terraforming would have to undo.
So start with the evidence, because it's overwhelming, and it took decades of spacecraft to assemble. As Jatan Mehta wrote for The Planetary Society in 2021, Mars was once an Earth-like world. When life was just getting started on our own planet — somewhere between three and a half and four billion years ago — Mars had lakes of liquid water and probably flowing rivers, a thicker atmosphere, a magnetic field shielding it from radiation, and the organic molecules that life is built from. That's not a poetic flourish. It's the read on the rocks. Orbiters mapped the river valleys. Rovers drove across the old lakebeds and drilled into the clays. The picture they built up is of a young Mars that, for a stretch of its history, looked a lot more like home.
Here's the part that reframes everything. Mars didn't lose that climate to some freak catastrophe. It lost it to physics that's still running right now, today, while you listen to this. And the trigger was something happening deep inside the planet.
To see why, you have to start with the magnetic field — and this is where most people's intuition goes sideways. The obvious assumption is that a magnetic field matters because of radiation, that it's mainly a shield for living things. That's true, but it's not the part that mattered for the climate. The deeper job a magnetic field does is protect the air. Picture the field as an invisible umbrella standing between the planet and the Sun. The Sun isn't just sending light. It's blowing a constant stream of charged particles outward in every direction — a wind of them. On Earth, that umbrella deflects the wind around us, the way a boulder splits a river. Take the umbrella away, and the wind hits the top of your atmosphere directly, and it starts blowing the air off into space, molecule by molecule.
That's what happened to Mars. Sometime between three and four billion years ago, the planet's core cooled enough that its global magnetic field shut down. The umbrella closed. And the solar wind — that incessant stream of energetic particles from the Sun, as Mehta put it — started stripping the atmosphere and the surface water away. Over billions of years, it scoured most of it off.
Now, you might reasonably ask — how could anyone possibly know that? It's a story about events four billion years gone. The answer is one of the genuinely satisfying detective stories in planetary science, and it has a name: MAVEN. That stands for Mars Atmosphere and Volatile Evolution, a NASA spacecraft that's been orbiting Mars since 2014, built for exactly one job — to measure how fast Mars is losing its air to space right now. And here's the elegant part. You can't film four billion years of history. But you can watch the leak that's still happening, measure its rate, measure how the solar wind drives it, and run the clock backward. MAVEN caught the planet in the act. It measured atmosphere escaping to space and tied that loss directly to the solar wind hammering an unprotected planet. That present-day measurement is the evidence for the ancient story. The leak you can see today is the same leak that drained the planet.
Stay with that for one more step, because it's the hinge of the entire course. The thing that killed Martian habitability wasn't a one-time event you could reverse with a single fix. It was a process — slow, relentless, and still going. Mehta put the loss in a single sentence that's worth holding onto: losing the magnetic field let the solar wind strip away most of the planet's atmosphere and surface water, turning Mars into the chilly desert we see today. The chill and the dryness aren't two separate problems. They're the same wound. Thin air means no greenhouse blanket, which means brutal cold, which means no liquid water can survive. Pull the air off a planet, and everything else follows.
So where did it all go? Some of it — a lot of it — is simply gone, blown into space over eons. But not all of it. And this is the hopeful half of the inventory, because Mars didn't lose every drop of its water. A huge amount of it froze in place and stayed.
You can see some of it from orbit with the naked eye, in a sense — the bright polar caps at the top and bottom of the planet, layered ice that grows and shrinks with the Martian seasons. But the polar caps are the obvious part. The surprising part is buried. Beneath the surface, across enormous stretches of Mars, there's water ice — frozen, hidden, waiting. And we know it's there because of a clever instrument with an even cleverer trick.
The instrument is called MARSIS — Mars Advanced Radar for Subsurface and Ionosphere Sounding — flying on the European Space Agency's Mars Express orbiter. Here's how it works, in kitchen-table terms. It fires radio waves down at the planet. Most bounce off the surface, but some punch through the dirt and reflect off whatever's underneath. Different materials send back different echoes. Rock answers one way. Ice answers another. By reading those echoes, you can map what's hidden under the ground without ever digging — like a doctor reading an ultrasound, except the patient is a planet and the probe is a radio wave from orbit. What the radar surveys found is that Mars holds vast reserves of subsurface water ice, locked away in the crust where the cold preserves it.
That matters enormously for the chapters ahead, so file it away: Mars is not actually short on water. The water mostly didn't escape. It froze and went underground. The thing Mars catastrophically lost was its atmosphere — the gas that kept it warm and kept that water liquid.
Which brings up the question this whole chapter has been circling, and it's worth pausing on. If the water's still mostly here, and we know exactly what went wrong — the air leaked away after the magnetic field died — then couldn't we just put it back? Warm it up, melt the ice, rebuild the atmosphere, undo the damage? … Hold that thought, because the honest answer is the spine of everything that follows, and it splits into two pieces that pull in opposite directions.
The first piece is encouraging. Knowing the failure mode is half the battle. A doctor who understands exactly how a disease progresses is in a far better position than one staring at a mystery. We know the patient's history. We know the wound. The river valleys and the buried ice tell us the raw materials existed and partly remain.
The second piece is the sobering one, and it's where this chapter quietly hands off to the next. The same physics that drained Mars the first time never stopped. There's still no global magnetic field. The solar wind is still blowing. So even if you could somehow rebuild a thick atmosphere tomorrow, you'd be pouring air into a planet that's still leaking — slowly, but forever. Nature didn't just run the experiment in the wrong direction. It's still running it. That's not a reason to give up, but it changes what success would even mean.
The cleanest way to say it is this: Mars isn't a planet that never had a climate. It's a planet that had one and lost it — which means terraforming wouldn't be invention, it would be reversal. And reversal raises a brutal, specific question. To rebuild that lost atmosphere, you'd need raw material — carbon dioxide, mostly, the gas that does the warming. Is there enough of it still on Mars to actually do the job? Because in 2018, a team of NASA scientists sat down with twenty years of spacecraft data and added it all up — and the number they got changes the whole conversation.