Posted by An engineer asked how can we live on Mars will not answer with destiny. You get a parts list. Pressure vessels, oxygen plants, buried ice, kilowatts, spare seals, and a radio link that needs a quarter of an hour to say yes. Mars does not kill people dramatically. It kills them through a failed valve at three in the morning, the nearest specialist twenty light-minutes away.
The numbers that set the problem are not negotiable. Surface pressure averages about 610 pascals, roughly six-tenths of one percent of sea level on Earth. The air is around 95 percent carbon dioxide. Mean surface temperature sits near minus 60 Celsius. None of that can be fixed; it can only be enclosed, powered and maintained, which makes settlement a question of hardware.
The Lunchbox That Made Air Out of the Sky
On 20 April 2021 a device the size of a car battery, bolted inside the Perseverance rover, pulled Martian air through a filter, heated it to about 800 degrees Celsius and split the carbon dioxide. It made 5.4 grams of oxygen — roughly ten minutes of breathing for one person. MOXIE, the Mars Oxygen In-Situ Resource Utilization Experiment, ran sixteen times before being retired in 2023: daylight and darkness, different seasons, thin air after the dust had settled. Total yield, about 122 grams.
A trivial amount of gas, and an enormous result. Michael Hecht, the principal investigator at MIT’s Haystack Observatory, spent years arguing that the first thing you send to Mars is a machine that makes what you cannot afford to ship. Breathing is not even the main cost — getting a rocket back off the surface takes tons of oxidiser.
MOXIE was roughly a two-hundredth of the scale a real plant needs, and it proved the ugly part too: solid-oxide electrolysis at 800 degrees eats electricity. Scaling it means scaling the power plant first.
Digging for Water That Is Already There
In June 2008 the Phoenix lander scraped a shallow trench at 68 degrees north and exposed a few bright white crumbs. Over the next four sols they disappeared. Not dissolved, not blown away — sublimated, which is what water ice does in Martian air. It was the first time anyone had touched water on another planet.
The mapping has improved since. Mars Odyssey’s neutron spectrometer found hydrogen across huge swathes of the mid-latitudes in 2002, and NASA’s Subsurface Water Ice Mapping effort has stitched radar and thermal data into charts of where ice sits shallow enough for a drill rather than a mine. Arcadia Planitia keeps coming up.
Water is the pivot everything turns on. Drink it, grow food in it, stack it around the bunks as shielding — then split it. Electrolysis gives oxygen for the lungs and hydrogen for chemistry, and hydrogen fed into Martian carbon dioxide over a nickel catalyst yields methane and more water. That is the Sabatier reaction, hardly exotic: such a unit has run aboard the International Space Station since 2010. Methane plus oxygen is rocket fuel. Hence landing where the ice is, not where the geology is pretty.
The First Martian Houses Will Be Buried
Curiosity’s radiation detector has logged the surface dose since 2012: around two-thirds of a millisievert a day, call it a hundred times what you accumulate standing in a field on Earth. Mars has no global magnetic field and almost no atmospheric column, so cosmic rays arrive unimpeded, and a solar particle event can deliver in hours what the background delivers in months.
The only shielding that works is mass, and mass is the one thing Mars has in abundance. Which is why the practical answer to how can we live on Mars mostly points downward.
The habitat gets covered: a few metres of packed regolith over an inflatable shell, or a module lowered into a trench and backfilled, or a structure printed from local material. NASA’s 3D-Printed Habitat Challenge spent four years on that last idea, and its 2019 winner, AI SpaceFactory’s MARSHA, was a tall printed shell of basalt fibre and bioplastic rather than a dome — a vertical cylinder handles internal pressure better.
Then there are the lava tubes. Mars Odyssey’s thermal camera spotted a cluster of dark pits on the flanks of Arsia Mons in 2007 — skylights, almost certainly, where an old lava conduit’s roof collapsed. Lower gravity means Martian tubes should dwarf anything on Earth, potentially hundreds of metres across. Such a cave solves radiation, micrometeorites and thermal swing at once. It also means building in the dark, somewhere nobody has surveyed, with no idea how stable the roof is after three billion years.
The Suit Problem
Body fluids boil at body temperature below about 6.3 kilopascals. Mars sits an order of magnitude under that line, so a suit is not clothing, it is the smallest spacecraft anyone will ever build. Current station suits run at 4.3 psi of pure oxygen, low enough that crews must pre-breathe for hours before every walk to avoid the bends. Nobody does that twice a day for thirty years.
The alternatives are unfinished. Higher pressure means stiffer joints and exhausting work. Mechanical counterpressure — the skintight approach Dava Newman’s group at MIT has developed — squeezes the body directly instead of inflating a balloon around it, lighter and far more mobile, if anyone solves getting into the thing.
And there is the dust. Martian fines are electrostatic, sharp-edged and laced with perchlorates, which Phoenix measured at close to one percent of the soil at its site and which are toxic. Apollo crews tracked lunar dust into the lander and it wrecked seals and irritated lungs within days. The favoured fix is the suitport: the suit hangs permanently outside the airlock and you climb in through the back, so the dirty surface never enters the room where people sleep.
Salad, and What Happened at Biosphere 2
Astronauts have eaten space-grown food since 10 August 2015, when the Veggie unit on the station produced red romaine the crew was cleared to eat. Chile peppers followed in 2021. On the ground, Wieger Wamelink’s group at Wageningen has spent a decade growing tomatoes, rye and radishes in NASA’s Mars simulants, and the crops come up — but simulant is ground volcanic rock, not real regolith with its perchlorates and its total absence of anything biological.
Martian farming will be hydroponic and artificially lit, at least at first.
The cautionary tale is still Biosphere 2. Eight people sealed themselves into a glass ecosystem in the Arizona desert in September 1991, and over sixteen months the oxygen inside quietly fell from 20.9 percent to 14.2 — the equivalent of living at 4,000 metres. It had to be pumped in from outside. The culprit was curing concrete absorbing carbon dioxide and soil microbes eating oxygen faster than plants replaced it: a bookkeeping error in a closed loop, found only because people were inside it. Nobody has yet run a fully closed life-support system for long. The station recycles roughly 98 percent of its water and is resupplied constantly.
Panels, Dust, and a Reactor the Size of a Bin
Mars gets about 43 percent of the sunlight Earth does, and the sky is dusty even on a good day. Solar works — it has powered rovers for two decades — but its failure mode has a name and a date. Opportunity fell silent in June 2018 when a planet-encircling dust storm blotted out the sun, and InSight’s mission ended in December 2022 with its arrays smothered in dust nobody could brush away. A rover going quiet is a scientific loss. A settlement going quiet is a body count, since heaters, oxygen and pressure control draw from the same bus.
Which is why NASA has been quietly rebuilding a technology it shelved decades ago. The Kilopower project ran a full-power test of its KRUSTY reactor in Nevada in March 2018: a solid uranium-235 core about the size of a paper towel roll, sodium heat pipes carrying heat to Stirling engines, no moving parts in the reactor itself. One kilowatt electric in the demonstration, scalable by design to ten. Four such units would run an outpost through a dust storm, through winter, through the night. Whether anyone accepts launching fission reactors to another planet is a separate question, and not a technical one.
Most serious designs hedge and take both, plus batteries.
Twenty Minutes Is a Long Time
Depending on where the two planets sit in their orbits, a radio signal takes between roughly four and twenty-four minutes to cross the gap one way. Double it for an answer. And every 26 months Mars slides behind the Sun and controllers stop sending commands entirely for about two weeks, because the corona corrupts the link.
This is the part people underestimate when they ask how can we live on Mars. There is no mission control in any real sense — only advice, arriving late. A settlement holds its own diagnostics, its own spare parts, its own surgeon and its own machinist, and it decides what to do about a depressurising module before Houston has finished hearing about it. The delay is what turns a base into a community, whether anyone plans for it or not.
None of this is impossible. Most of it is unfinished, which is a different thing: working demonstrations of perhaps half the pieces, paper studies of the rest. If you want to follow the hardware as it actually gets built — oxygen plants, ice maps, reactors, suits — SETIworld covers this beat, and the argument is better with more people in it.
