Posted by Mars today carries an atmosphere of about six millibars, six thousandths of the pressure at sea level on Earth, and it is almost all carbon dioxide. Step outside without a suit and the water in your tissues would start to boil at body temperature. The reason anyone still asks what happened to Mars is that the same world, four billion years ago, had rivers running across it.
Not maybe-rivers. Branching valley networks cut into the southern highlands, deltas with tilted foreset beds, crater floors filled with the fine mudstone that only settles out of standing water. The case for a wet early Mars is about as solid as planetary geology gets. The hard part is the mechanism: not that Mars was wet, but why it stopped.
The short answer: a magnetic field that switched off, a star that peeled the air away particle by particle, and a crust that quietly drank a great deal of what was left. No one of the three is enough on its own.
The Dynamo Went Out Early
Mars Global Surveyor arrived in September 1997 with a magnetometer built by a team under Mario Acuña, and on long aerobraking passes it dipped low enough to feel what the crust was doing. There was no global field. Earth’s, churned up by liquid iron in the outer core, wraps the planet in a shield. Mars had nothing like it.
What it found instead was stranger and more useful: stripes of magnetized rock across the ancient southern highlands, in Terra Cimmeria and Terra Sirenum, alternating in polarity, running for hundreds of kilometres, magnetized more strongly than anything comparable in Earth’s crust. Rock locks in whatever field surrounds it as it cools past a certain temperature, and then holds that record.
Those stripes are a fossil of a magnetic field that no longer exists.
The dating argument is the elegant part. The huge impact basins, Hellas and Argyre and Isidis, are magnetically blank. The impacts that dug them cooked the crust past its recording threshold, and when that rock cooled there was no field left to imprint on it. Those basins formed roughly four billion years ago, so the dynamo ran while the oldest highlands were laid down and had stopped by the time the basins were punched out. Early. Why it quit is still argued over, though Mars being about a ninth of Earth’s mass is the usual starting point.
What MAVEN Watched Happening
MAVEN launched in November 2013 and reached orbit the following September with a narrow brief: measure how fast Mars is losing its atmosphere now, and by which routes.
There turned out to be several. Solar wind ions slam into the upper atmosphere and knock neutral atoms off it, a process called sputtering. Ultraviolet light strips electrons from atmospheric molecules, and the electric field carried by the solar wind picks up the resulting ions and hauls them off. Oxygen ions recombine with electrons and the atoms fly apart fast enough to escape outright. In 2015 the mission put a figure on the total: on the order of a hundred grams per second. Trivial, until you multiply by four billion years. During solar storms the rate jumps tenfold and more.
The young Sun was rougher anyway: faster rotation, a denser wind, extreme ultraviolet at levels today’s Sun cannot manage.
MAVEN’s neatest result came from argon. Argon reacts with nothing, so it cannot hide inside minerals or dissolve into an ocean. It comes in a light isotope and a heavy one, and essentially the only process that preferentially strips the light one off the top of the atmosphere is sputtering. Comparing the ratio high up against the ratio near the surface, the team concluded in 2017 that around two thirds of the planet’s argon has been thrown into space. If two thirds of the argon went that way, the bulk of the carbon dioxide almost certainly did too, and the pressure that keeps liquid water stable went with it.
Counting the Missing Ocean in Deuterium
Hydrogen comes in a heavy version, deuterium, carrying a neutron. At the very top of an atmosphere, where atoms escape one at a time, the lighter kind gets away more easily, so water that has been leaking for billions of years leaves behind water enriched in deuterium. The ratio works as a counter for what has gone.
In 2015 a team led by Geronimo Villanueva mapped that ratio from the ground, using infrared spectrographs on Keck II in Hawaii and the Very Large Telescope in Chile to pick heavy water out from ordinary water region by region. The enrichment came out at roughly seven times the value in Earth’s ocean. Run backwards, that implies Mars has lost something like 87 percent of the water it started with.
Put the missing share back and you get, at minimum, a layer 137 metres deep spread evenly over the globe. Mars is not even. The northern lowlands sit kilometres below the southern highlands, so a likelier arrangement pools it into a northern ocean covering perhaps a fifth of the surface. The 87 percent is robust. The ocean is interpretation.
The Clays, Jezero and Gale
Mineralogy caught the collapse in progress. OMEGA on Mars Express, and later CRISM aboard Mars Reconnaissance Orbiter, mapped clay minerals across the oldest terrain on the planet. Clays do not form dry. They need water working on rock for long stretches, in conditions that are not strongly acidic. In younger ground they give way to sulfates, which want acid and far less water, and younger still come the iron oxides that give the modern surface its colour.
Then the rovers went to look up close.
Curiosity landed in Gale Crater in August 2012 and within months was drilling mudstone at Yellowknife Bay that had settled out of a standing lake, fine-grained and near neutral in pH, the sort of water you could drink. Perseverance went to Jezero Crater in February 2021 because Jezero has a delta, and deltas trap the fine sediment that preserves delicate things. It found clays there and carbonate-bearing rock around the crater margin, and has been sealing cores into tubes ever since. Whether those tubes reach a laboratory on Earth is a budget question, not a geological one.
The Water That Went Into the Rocks
Escape to space is not the whole ledger, and this is where the popular account of what happened to Mars usually goes wrong. Water also reacts with basalt, binding into the structure of clays and other hydrated minerals and staying there. On Earth that is temporary, because plate tectonics drags crust into the mantle and cooks the water back out. Mars has no plate tectonics. What goes into the rock stays in the rock.
In 2021 Eva Scheller and colleagues at Caltech tried to balance the books, combining escape rates, the deuterium record and the crust’s capacity to hold water in minerals. Their conclusion: between 30 and 99 percent of the planet’s original water sits in the crust, not in space. The range is embarrassingly wide, which is honest of them, and a genuinely different answer from the one about the solar wind.
Carbon dioxide has a parallel story. Curiosity’s drill turned up siderite, an iron carbonate, in the sulfate-rich layers of Mount Sharp, reported in 2025 in quantities large enough to suggest carbonate formation pulled a real share of the atmosphere into stone. Rock breathing the air in and never giving it back.
Plenty of water is also simply still there, frozen. Mars Odyssey’s neutron spectrometer detected hydrogen in the top metre of soil across the high latitudes in 2002, and radar sounders have since traced buried ice sheets, one under Utopia Planitia holding about as much water as Lake Superior. In 2024 a group working with InSight’s seismic data argued for something stranger: liquid water in fractured rock 11 to 20 kilometres down, potentially enough to cover the planet a kilometre deep. That is an inference from wave speeds, not a drill hole.
Why the Cause Matters More Than the Date
Mars did not die on a particular afternoon. The valley networks mostly stop being carved around 3.7 billion years ago, yet water kept turning up afterwards in bursts, in the outflow channels and in ice that advanced and retreated with the planet’s wobbling tilt. What changed was the trend, not a switch. A world can stay habitable in patches long after it has stopped being habitable as a whole.
Which is the part worth carrying to other stars. Earth and Mars formed out of much the same material, both got volcanoes and impacts and water in contact with rock, and only one still has an ocean. The difference came down to mass, internal heat, how long the dynamo lasted. When a survey turns up a rocky planet at the right distance from its star, that distance is the easy number and the least informative one. Mars sits inside the outer edge of the Sun’s habitable zone right now, and it is a desert.
Asking what happened to Mars is really asking how long any planet gets to stay comfortable. We have one worked example next door, with the evidence still lying on the surface because there is no rain to wash it away and no tectonics to swallow it. If that is the kind of problem you enjoy chewing on, SETIworld follows this work as it arrives, from the cores waiting in Jezero to the atmospheres being teased out of planets around other stars, and there is room in the argument for anyone who wants to push back.
