Mars habitat design — pre-deployed, ISRU-fed, perchlorate-aware

Every credible Mars human architecture since the early 1990s has been built around the same observation: you cannot fly the habitat and the crew on the same launch window. The trip is too long (6-9 months one-way) and the surface stay too long (typically 500 days, dictated by orbital mechanics — the next Mars-to-Earth Hohmann window is 500 days after arrival) for the architecture to fit on a single launch. Solution: the habitat lands first, on a launch window 26 months before the crew, fully autonomous. By the time the crew arrives, the habitat has been operating for 26 months, the ISRU plant has been making oxygen and methane from the atmosphere and from subsurface ice for 18 months, the return vehicle is fully fuelled and pre-tested. NASA's Design Reference Architecture 5.0 (2009) codified this; SpaceX's Starship Mars architecture inherits it; every Mars architecture variation since then has retained the pre-deployment pattern.

Radiation shielding is harder than on the Moon. Mars has a thin (6 mbar) atmosphere that absorbs some galactic cosmic ray secondary cascades but no magnetic field. Surface dose: ~0.6 mSv/day, very similar to the Moon. Mitigations are the same family: regolith piled around or over the habitat (1-2 m of regolith blocks the bulk of the dose), or burial — Mars Design Reference Architecture 5.0 baselines pre-deployed habitat modules with 30-50 cm of regolith hand-piled by crew + robotic excavator. Subsurface lava tubes (detected at Pavonis Mons, Hadriacus Mons, and other volcanic provinces) are the long-term shielding target, similar to the lunar plan. Water-ice walls also feature in some architectures — water is roughly as effective per kilogram as regolith for shielding, with the bonus that you'd be hauling water anyway.

Dust on Mars is fundamentally different from lunar dust. Lunar dust is mechanical — micrometeorite-fragmented, electrostatically charged, glass-sharp. Mars dust is chemical: regolith samples returned by the Phoenix lander (2008) and confirmed by Curiosity and Perseverance contain 0.5-1% by mass of perchlorate (ClO₄⁻) salts. Perchlorates are thyroid toxins; at the surface concentration of ~0.5%, breathing Mars dust unfiltered would deliver a daily dose orders of magnitude above the recommended human limit within minutes. Architectures plan for HEPA-class filtration on every airlock, regolith-contact protocols requiring decontamination shower before re-entering pressurised volume, and ECLSS designed to detect perchlorate ingress in trace amounts. Mars dust also causes the dust storms — global storms (occurring ~every 3 Mars years, the most recent in 2018, which ended the Opportunity rover) blanket the planet in suspended dust for weeks, dropping solar panel output to ~10%. Crewed missions cannot run on solar alone through a global dust storm; baseline architectures include a 10-40 kWe fission reactor (Kilopower-derived, KRUSTY ground-tested in 2017-2018).

ISRU is non-negotiable for Mars. The mass ratio for return-to-Earth makes it impossible to ship return fuel from Earth; you must make it on-site. MOXIE (Mars OXygen In-situ resource utilization Experiment) flew on Perseverance and demonstrated O₂ production from Martian atmospheric CO₂ at small scale (5-12 g/hour, totalling 122 g across 16 runs from 2021-2023). Scale-up factor for a crewed mission: ~200×, producing 2-3 kg/hour for 18 months, total 25-30 tonnes of O₂ — enough for ascent vehicle oxidiser + crew breathing margin. Methane for ascent vehicle fuel comes from the Sabatier reaction: CO₂ + 4 H₂ → CH₄ + 2 H₂O, with H₂ sourced from electrolysed water mined from subsurface ice deposits (detected widely by MRO SHARAD radar). The whole ISRU plant is part of the pre-deployed payload, runs autonomously for 18-26 months before crew arrival, and produces test propellant samples that crewed mission can verify upon arrival before committing to the ascent burn.