Terraforming

Terraforming means deliberately changing a planet's environment until unprotected humans, or the biosphere they depend on, can live on its surface. In practice the word means Mars, and the framing under test is that Mars can be made habitable.

The decisive constraint is inventory, and that is what makes this subject unusual. Most speculative engineering fails on cost, on scale, or on a physical result nobody has produced. This one fails on a measurement: somebody went and counted the carbon dioxide. Jakosky and Edwards put everything mobilisable with present or near-future technology at about 0.020 bar against the roughly 1 bar needed — a factor of about fifty — delivering less than 10 K of warming against the roughly 60 K required. Their sentence is the shortest summary of the field: terraforming Mars is not possible in the foreseeable future using the CO₂ resources available on the planet.

Everything else here sits downstream of that. The 2024 nanoparticle result is real and is not terraforming; the aerogel work is the part that survives and is regional; the ethical literature has six named arguments and a philosopher's ranking of them; and the timescales in circulation span seven orders of magnitude because different authors are answering different questions. The boundary with Artificial Magnetospheres is one sentence — that brief owns loss, this one owns inventory, and stopping atmospheric loss is not the same as having an atmosphere — and this page states plainly that on the measured numbers, loss is not what binds.

Established Jakosky and Edwards, in Nature Astronomy in 2018, measured the reservoir and the answer is no. The requirement is about 1 bar of CO₂, roughly 2,500 g per square centimetre, to produce the ~60 K of warming liquid water needs. The present atmosphere is 6 mbar, about 15 g per square centimetre. Their reservoir-by-reservoir breakdown, in units of the surface pressure each would produce: polar ice caps ~12 mbar, accessible; adsorbed in regolith under 40 mbar, extremely difficult; near-surface carbonates 12–150 mbar, limited access; polar clathrates under 150 mbar, inaccessible; deep carbonates about 1 bar, inaccessible; and more than 0.5 bar already lost to space, permanently gone.

Established The total mobilisable with present or near-future technology is about 0.020 bar — 20 mbar — and it delivers “less than 10 K.” Established Derived from those published figures, the shortfall is a factor of about 50 in pressure and about 6 in warming; the pressure ratio is the one that communicates. Their conclusion, verbatim: “Terraforming Mars is therefore not possible in the foreseeable future by utilizing CO₂ resources available on the planet.”

Established The reasons the big reservoirs cannot be used are physical and specific. Carbonates decrepitate at about 300 °C, so releasing them requires industrial processing of planetary volumes of rock. Adsorbed CO₂ is diffusely distributed, never fully releases at thermal equilibrium, and a temperature increase takes about 10⁴ years to penetrate a hundred-metre regolith column. These are not engineering difficulties in the ordinary sense; they are statements about where the carbon is and how slowly heat travels through rock.

Frontier An independent 2026 analysis reaches the same verdict from a different direction, and the convergence is worth stating explicitly. Turyshev's Terraforming Mars: Mass, forcing, and industrial throughput constraints re-derives the problem as a systems-constraint exercise. His figures: 3.89 × 10¹⁵ kg of atmosphere per millibar of Martian global-mean surface pressure; human-relevant pressures needing gas inventories at the 10¹⁷–10¹⁸ kg scale; a breathable mix requiring 9.0 × 10¹⁷ kg of O₂ for 21 kPa and 1.9 × 10¹⁸ kg of N₂ buffer for 50 kPa, with the reversible work of water electrolysis for the oxygen alone at about 1.3 × 10²⁵ J. Target mean surface temperatures of 250–273 K need longwave optical depth of order unity and about 10² W/m² of direct absorbed-solar forcing, implying 10¹³–10¹⁴ m² of orbital reflector area if done that way.

Frontier On the endogenous route Turyshev reproduces Jakosky and Edwards exactly: a representative 20 mbar case yields 10 K or less. His conclusion is that “no surveyed abiotic global open-atmosphere pathway simultaneously closes pressure, composition, warming, power, throughput, retention, sink, and operations constraints,” and that regional and enclosed habitability is the physically feasible staged approach. Established Two independent analyses, eight years apart, by different methods, on the same verdict.

Frontier The most interesting recent development is the 2024 nanoparticle result, and it is routinely over-reported, so the precision matters. Ansari, Kite, Ramirez, Steele and Mohseni, in Science Advances in August 2024, propose engineered conductive nanorods of aluminium or iron, about 9 micrometres long with a 0.16 by 0.16 micrometre cross-section — an aspect ratio near 60 to 1, with a more mass-efficient iron variant at 7.5 by 0.08 micrometres. The required column loading is 160 mg/m² of aluminium nanorods. The claimed efficiency is “>5000× more effective, on a warming-per-unit-mass-in-the-atmosphere basis, than the current state of the art” — the comparison being about 1.1 kg/m² of an optimal greenhouse-gas mix, mostly C₃F₈, against about 160 mg/m² of nanorods.

Frontier What they claim it does. Warming of 30 K or more globally, with 40–50 K warm-season increases in the southern hemisphere, enough for summertime liquid water where shallow ground ice exists. Pressure rises about 20% within months as CO₂ ice sublimes, with a further factor of 2 to 20 over a period “that could be as long as centuries.” Nanorods settle more than ten times more slowly than natural Mars dust, with an assumed particle lifetime of 10 years against a Brownian coagulation timescale of about six.

Established And what they explicitly do not claim, in their own words, twice. “Raising Mars’ temperature, by itself, is not sufficient to make the planet’s surface habitable for oxygenic photosynthetic life: barriers remain,” and “Nanoparticle warming, by itself, is not sufficient to make the planet’s surface habitable again.” The named remaining barriers are about 300 ppmw nitrates in the soil, “very little O₂” in the air, and perchlorate-rich soil requiring biological remediation. They raise a risk against themselves: centuries of excessive warming could drain ground water into empty pore space if the ice is underlain by voids. And they state that “effective particle lifetime remains a major uncertainty in our model.”

Frontier There is an arithmetic tension in that paper and this page prints it rather than resolving it, because resolving it would require the full text and a calculation nobody has published. A standing inventory of 160 mg/m² across Mars's 1.448 × 10¹⁴ m² is about 2.3 × 10⁷ kg, roughly 23,000 tonnes of aluminium, or about 8,600 m³ of metal. Replaced on the paper's own nominal 10-year particle lifetime, that is about 860 m³ per year. But the paper's stated operational requirement is particle fountaining at about 30 litres per second, roughly 7 × 10⁵ m³ per year of metal production — described as “equivalent to a mine with half-width 350 m lengthening 250 m annually.” Speculative Those two numbers differ by about 800 times. They reconcile only if the effective residence time in the operating model is days rather than the nominal ten years — that is, if most released mass is lost promptly. The derivation is this brief's, both inputs are the paper's, and the tension is stated as an open question rather than as a refutation.

Established The consequence for how the headline should be quoted is exact. The >5000× figure is per unit mass in the atmosphere, not per unit mass produced, and the production requirement is set by the fountaining rate rather than by the column loading. Printing the 5000× figure without the 30 litres per second is the exact over-simplification this page exists to prevent.

Frontier The part of the field that survives contact with the inventory result is regional, and it has measurements behind it. Wordsworth, Kerber and Cockell's silica-aerogel work uses 2–3 cm of aerogel, optimum about 2.5 cm, and measures temperature differences of more than 45 K under Mars-like conditions at 150 W/m², with tiles reaching more than 50 K at just 2 cm thickness. It works within the ±45-degree latitude band where year-round solar flux is high, attenuates UV-AB strongly and UV-C almost totally — 2 cm tiles cut UV-C transmission below 0.5%. The authors' own framing is that this is “much more achievable than global atmospheric modification,” avoiding “massive environmental modification that will be well beyond human capability for the foreseeable future.” Established The paper does not state a treatable area or an aerogel mass per square metre, and neither is invented here.

Established The founding paper made material availability the deciding variable, which is exactly what was later measured. McKay, Toon and Kasting's 1991 Making Mars habitable states that success “would depend on the abundance, distribution and form of materials on the planet that could provide carbon dioxide, water and nitrogen.” Frontier Twenty-seven years later Jakosky and Edwards measured that abundance and the answer was no. Only the abstract of the 1991 paper was obtained during research for this brief, so no timescale from it is quoted here.

Established The other inventories are no kinder. Integrated water loss from 4.2–3.5 Ga is equivalent to a 23-metre global equivalent layer. Water still present near the surface is mapped by NASA's SWIM project, whose fourth-iteration map of 26 October 2023 covers buried ice from the equator to 60 degrees north within the upper metre, using MRO, Mars Odyssey and Mars Global Surveyor data, fresh ice-exposing craters and polygonal terrain — with a distribution that is markedly non-uniform. The release gives no total inventory and no global equivalent layer; SWIM is a mission-siting product, not a water census, and this page does not treat it as one. Established Nitrogen is arguably the worse problem and is under-discussed: about 300 ppmw nitrates in the soil against Turyshev's buffer requirement of 1.9 × 10¹⁸ kg of N₂. There is no polar cap of nitrogen.

Established And the soil is actively hostile to the organisms the roadmap depends on. Phoenix measured 0.4–0.6 wt% perchlorate in Martian soil, with localised brine seeps possibly reaching several per cent. Wadsworth and Cockell showed that UV-irradiated perchlorate is bactericidal: Bacillus subtilis lost viability in 30 seconds at 0.6 wt% under UV against 60 seconds for UV alone; 5 wt% killed in 30 seconds; and perchlorate plus iron oxides plus hydrogen peroxide produced a 10.8-fold increase in cell death over UV alone. Their conclusion is that irradiated perchlorates “render the present-day surface more uninhabitable.” Frontier Two consequences follow and they point opposite ways: terraforming must remediate a soil chemically lethal to the photosynthetic microbes its roadmap requires — which Ansari et al. name themselves — and the same finding lowers forward-contamination risk, which cuts in favour of the interventionist side of the ethics argument.

Frontier Is there any pathway that closes? Turyshev's answer is no for global abiotic open-atmosphere routes, on eight simultaneous constraints — pressure, composition, warming, power, throughput, retention, sink, operations. The open question is whether a biological route, or a hybrid, closes constraints that no abiotic route can, and nobody has run that analysis. The community's own 2025 agenda names “fundamental physical, chemical and biological constraints” as the research priority, which is a concession that they are not currently characterised.

Frontier How much metal does nanoparticle warming actually cost? This is the live technical question in the newest result and it is quantitative. The standing atmospheric inventory implied by 160 mg/m² is about 23,000 tonnes; the stated fountaining requirement is 30 litres per second, about 7 × 10⁵ m³ a year. Speculative The ratio is roughly 800 and is unexplained in the material available here. If the effective residence time is days rather than years, the efficiency headline and the industrial requirement are describing different quantities, and only one of them is what a mission would have to build.

Speculative What is the effective particle lifetime? The authors name this themselves as “a major uncertainty in our model.” Nanorods settle more than ten times more slowly than natural dust; the assumed lifetime is ten years; the Brownian coagulation timescale is about six. Everything in the concept's economics scales inversely with this number, and it is not measured.

Frontier Would warming release ground water in the wrong direction? Ansari et al. raise the risk against their own proposal: centuries of excessive warming could drain ground water into empty pore space if shallow ice is underlain by voids. The distribution of subsurface voids beneath Martian ground ice is unmapped, and SWIM covers only the upper metre to 60 degrees north.

Frontier Does the regional route scale, and how far? The aerogel result is a measured 45–50 K temperature difference at 2–3 cm within ±45 degrees latitude, with UV-C transmission below 0.5%. The paper states no treatable area and no mass per square metre, so the step from a tile to a habitable region is unquantified in the primary literature — which is the single most useful calculation missing from the surviving branch of the field.

Speculative Is the biological step possible at all? The Sagan-to-present sequence is warming, then oxygenic photosynthesis by engineered microbes, then slow oxygen build-up, then more complex life. Handwave No step after the first has a quantified timescale anywhere in the retrieved literature, and the organisms would have to survive a perchlorate-rich, UV-irradiated soil that kills Bacillus subtilis in thirty seconds.

Frontier Is atmospheric loss a precondition at all? The standing framing says a magnetic shield must come first. Established The measured escape rate is 2–3 kg/s against an atmospheric mass of about 2.3 × 10¹⁶ kg, which is roughly 250 million years to empty, and two measurement groups now hold that magnetisation may not reduce escape and that Mars's ion escape is supply-limited. This page keeps the dependency on Artificial Magnetospheres and states plainly that on the numbers it is not what binds.

Speculative And the position at the edge of the hypothesis space, stated because leaving it out would be worse. A minority view holds that the inventory result is a statement about present technology only — that deep carbonates and clathrates are inaccessible in the way deep-sea oil was inaccessible in 1900. Handwave Nothing in the retrieved literature develops that argument quantitatively, and Jakosky and Edwards' own barrier is thermodynamic rather than mechanical: decrepitating carbonates at 300 °C across planetary volumes is an energy requirement, not a drilling one.

Established Inventory is the bottleneck, and it is the kind that does not yield to engineering. About 20 mbar mobilisable against about 1 bar needed. A better drill, a cheaper rocket or a fusion reactor does not create carbon dioxide that is not on the planet, and that is what distinguishes this from every other constraint in this brief.

Established Heat transport through regolith is the second, and it is measured in units nobody expects. A temperature increase takes about 10⁴ years to penetrate a hundred-metre regolith column, so the adsorbed reservoir — the largest of the merely-difficult ones — cannot be released on any project timescale even if the energy were available.

Frontier Industrial throughput is the bottleneck for the newest route. Sustaining nanoparticle warming requires about 7 × 10⁵ m³ of engineered metal per year, on the paper's own figure — a mine 700 m across lengthening 250 m annually, on Mars, indefinitely. The warming is cheap in atmospheric mass and expensive in production rate, and those are different quantities.

Established Nitrogen has no accessible reservoir at all. About 300 ppmw nitrates in soil against a buffer-gas requirement of 1.9 × 10¹⁸ kg. The CO₂ problem at least has polar caps; the nitrogen problem has nothing analogous, and it is discussed far less.

Established Soil chemistry blocks the biological step before it starts. Perchlorate at 0.4–0.6 wt% under UV kills Bacillus subtilis in thirty seconds, and the combination with iron oxides and peroxide multiplies cell death nearly eleven-fold. Remediation of a planetary surface is a prerequisite for a roadmap whose second step is engineered photosynthetic microbes.

Frontier Energy is a bottleneck at a scale worth stating. Turyshev puts the reversible work of electrolysing enough water for a breathable oxygen partial pressure at about 1.3 × 10²⁵ J — and that is the thermodynamic floor for one component of one constraint, before any inefficiency. Global human primary energy consumption is of order 10²¹ J a year.

Speculative And the bottleneck that is not a bottleneck: retention. It is treated as first in the popular ordering and it is last in the measured one. At 2–3 kg/s the present atmosphere empties in about 250 million years, so nothing about loss constrains a project that has not yet solved a fifty-fold inventory shortfall.

Established The adjudicated dependency is Artificial Magnetospheres, and this page keeps it while stating its weakness. That brief owns loss; this one owns inventory. The sentence both need: stopping atmospheric loss is not the same as having an atmosphere. Frontier On measured numbers the edge is narrative rather than binding — escape at 2–3 kg/s against 2.3 × 10¹⁶ kg gives about 250 million years, and Gunell (2018) and Ramstad and Barabash (2021) dispute that a field helps at all, with Mars's ion escape characterised as supply-limited. The shield is a retention precondition for a hypothetical thick atmosphere on geological timescales. The constraint that closes off the CO₂ route is Jakosky and Edwards.

Frontier The dependency that would actually change the verdict is a measurement of the deep reservoirs. Deep carbonates hold about 1 bar and are classified inaccessible; clathrates hold under 150 mbar and are classified inaccessible. An accessibility result — depth, distribution, energy of release — is the single scientific finding that would reopen the global route, and no mission has been proposed to obtain it.

Frontier A measured particle lifetime for engineered nanorods in the Martian atmosphere. The concept's own authors name it as their major uncertainty; the economics scale inversely with it; and it is the quantity that would resolve the 800-fold tension between column loading and fountaining rate set out above. It is measurable in a chamber before it is measurable on Mars.

Established A total water inventory, which does not currently exist in citable form. SWIM maps shallow ice to 60 degrees north within the upper metre and gives no total and no global equivalent layer; the 23 m figure in the literature is lost water, not present water. A bulk inventory would require a dedicated study this brief could not obtain.

Frontier An area-and-mass calculation for the aerogel route. The primary paper states neither a treatable area nor a mass per square metre. Since that route is the one the systems analysis identifies as physically feasible, its absent numbers are the field's most consequential missing arithmetic.

Established And two dependencies on publication access that bound this page. The perfluorocarbon baseline paper — Marinova, McKay and Hashimoto (2005) — returned 403 on both attempts, so nothing from it is quoted here beyond the ~1.1 kg/m² benchmark reported by Ansari et al. Sparrow's 1999 Environmental Ethics paper is paywalled and was not obtained, so his formulation appears here only at one remove, through Stoner and York, and is labelled as such wherever it is used.

Established The experiment that decided the field was an inventory audit, and it needed no new hardware. Jakosky and Edwards assembled existing spacecraft measurements into a reservoir-by-reservoir table. The single most consequential piece of work in terraforming research was a piece of accounting.

Established The best laboratory result in the field is the aerogel measurement. Tiles 2–3 cm thick, illuminated at 150 W/m² under Mars-like conditions, producing measured temperature differences above 45 K and cutting UV-C transmission below 0.5%. It is a bench measurement with a stated thickness and a stated flux, which is more than any global proposal has.

Frontier The decisive next experiment for the nanoparticle route is a chamber measurement of effective particle lifetime. Suspension behaviour, coagulation and settling for 9-micrometre conductive rods in a CO₂ atmosphere at Martian pressure and temperature. It is small, cheap, and would resolve both the authors' stated major uncertainty and the 800-fold arithmetic tension on this page.

Frontier A field trial of the regional route is the plausible flight experiment. A few square metres of aerogel over ground ice within the ±45-degree band, instrumented for temperature, UV and melt. It would test the only route the systems analysis calls feasible, at a scale a single lander could carry, and nobody has manifested one.

Frontier Perchlorate remediation at bench scale. Ansari et al. name perchlorate-rich soil as requiring biological remediation; Wadsworth and Cockell measured how lethal it is. Demonstrating a microbial or chemical route that survives 0.6 wt% perchlorate under UV would unblock the second step of every published roadmap, and it is a terrestrial laboratory problem.

Speculative And the experiment that would reopen the global question: a deep-reservoir accessibility probe. Drilling, sounding or seismic characterisation of deep carbonates. The reservoir large enough to matter is the one classified inaccessible, and nothing has been proposed to test whether that classification is permanent.

Established Requirement one, and it cannot be engineered: about 1 bar of CO₂ where about 0.020 bar is available. Every subsequent requirement is conditional on this one, and this one is a fact about the planet rather than about technology.

Frontier Requirement two, for the warming route that does exist: 160 mg/m² of engineered nanorods held in suspension planet-wide, sustained by production at about 30 litres per second. That is a mine of half-width 350 m lengthening 250 m a year, operating on Mars, indefinitely, on the paper's own description. Nothing about the fabrication of 9-micrometre conductive rods at that rate has been demonstrated anywhere, on any planet.

Frontier Requirement three, for the reflector route: 10¹³–10¹⁴ m² of orbital reflector area at Mars. That is two to twenty times the area of the terrestrial sunshade whose arithmetic is set out in Artificial Satellites for Climate Control — 20 million tonnes, 16 trillion elements, about 1,800 times all mass now in Earth orbit. The mass arithmetic is not repeated here; it is cited across.

Established Requirement four: energy at a scale with no terrestrial analogue. The reversible work of electrolysis for a breathable oxygen inventory alone is about 1.3 × 10²⁵ J, against global human primary energy use of order 10²¹ J a year. That is roughly ten thousand years of total human energy consumption, at thermodynamic perfection, for one component.

Frontier Requirement five, for the route that survives: aerogel manufacturing and emplacement. 2–3 cm of silica aerogel over an area nobody has specified, within ±45 degrees latitude, delivering 45–50 K and near-total UV-C attenuation. The engineering here is tractable and the missing number is the area, which the primary paper does not give.

Speculative Requirement six: a biology that works in the soil that exists. Engineered oxygenic photosynthesis is the second step of every roadmap, and the substrate kills its model organism in thirty seconds under UV at measured perchlorate concentrations. No published organism, remediation scheme or containment approach addresses this at planetary scale.

Established Artificial Magnetospheres is the adjudicated dependency and the seam is one sentence, printed in both briefs: stopping atmospheric loss is not the same as having an atmosphere. That page owns the escape physics, the 250-million-year timescale and the dispute about whether a field helps; this page owns what there is to retain. Both should say that on the measured numbers the edge is narrative rather than binding.

Frontier Artificial Satellites for Climate Control shares the toolkit at a different planet. Turyshev's reflector route needs 10¹³–10¹⁴ m² at Mars, two to twenty times Angel's terrestrial sunshade area. That brief owns the L1 hardware arithmetic and the terrestrial case; this one owns the Martian energy budget, and the mass numbers are cited rather than duplicated.

Established Mars Colonization owns the near-term work, and Turyshev's conclusion formally hands it over. The distinction is sealed habitats versus changing the planet. If regional and enclosed habitability is the physically feasible staged approach, then the aerogel line of research and the habitat programme are the live parts of this subject and global terraforming is not.

Established No edge to Climate Engineering, and this page resists one. That brief owns deliberate intervention in Earth's climate, its governance and its moral hazard. Terraforming ethics is a different literature with different named holders and a different core question — the intrinsic value of a lifeless or microbially inhabited world, not consent among affected populations. The phrase “planetary engineering” appearing in both is a coincidence of vocabulary.

Established Space Law and Governance owns the legal mechanisms — who decides, planetary-protection instruments, COSPAR policy. This brief owns the ethical arguments and names six of them with their holders. Named that way, the seam stays clean.

Frontier Nanomaterials manufacturing is the adjacent technology the newest route depends on entirely. Conductive rods 9 micrometres long with a 0.16-micrometre cross-section, produced at 7 × 10⁵ m³ a year from Martian feedstock. That is an in-situ industrial capability, not a climate technology, and it is developed elsewhere for other reasons.

Frontier The field's own community published a research agenda in 2025, and it should be read as what it is. DeBenedictis, Kite, Wordsworth and nine co-authors argue in Nature Astronomy that terraforming “has received surprisingly little rigorous study,” that progress in Mars science, climate science, launch and bioscience motivates a fresh look, and that “New techniques have emerged that could raise Mars’s average global temperature by tens of degrees within a few decades.” These are the authors proposing the research programme they would run — an interested party.

Frontier But two of its concessions run against that interest and therefore carry weight. It concedes that the field's rigorous study is thin, and it accepts the comparison “relative to the alternative of leaving Mars as a pristine wilderness” as a live one rather than dismissing it. An advocacy document that names the strongest objection to itself as a legitimate alternative is more useful than one that does not.

Established The institutional asymmetry in this subject is that the decisive results came from agency science rather than from the terraforming community. Jakosky is MAVEN's principal investigator; the inventory paper is NASA-sponsored work that closes off a popular ambition. An agency scientist publishing a result that forecloses a mission concept is interest running against the finding, and it is the strongest evidence on this page.

Frontier There is no terraforming programme anywhere, and that is a stable institutional fact rather than an oversight. No agency has a line item; the funded work is Mars science, habitat engineering and planetary protection. What exists is a literature, and it is small enough that a single 2018 paper reorganised it.

Established Planetary protection is the institutional regime that actually binds today. COSPAR policy and the forward-contamination rules constrain what any mission may do to the Martian surface, and they were written for sample return and landers rather than for deliberate environmental modification. The instruments belong to Space Law and Governance; what matters here is that they exist and were not designed for this.

Frontier And a publication-access fact that shapes what can be said. The perfluorocarbon baseline paper returned 403 twice and Sparrow's founding ethics paper is paywalled. A subject whose warming baseline and whose most-quoted ethical formulation are both unreachable is one where careful writing means citing at one remove and saying so.

Established The ethical debate has structure, six named arguments and a philosopher's ranking of them, and it deserves presenting as a live argument rather than a list of worries. Stoner's The Ethics of Terraforming: A Critical Survey of Six Arguments (2022) is the best available map, and the ratings below are his.

Established The three preservationist arguments, against. One, the duty of scientific conservation: in Stoner's formulation, “methods of investigation of an object of scientific interest are morally permissible only when they do not foreclose the possibility of future investigations.” He rates it strong. Two, the duty of wilderness preservation, resting on Holmes Rolston III's three criteria, with Valles Marineris and Gale Crater as extreme, historically significant and aesthetically valuable. Rated strong. Three, avoiding colonialist vices, held by Robert Sparrow and Keekok Lee: terraforming rhetoric exhibits arrogance and insensitivity to distinctive beauty. Rated compelling — the highest rating he gives any argument on either side.

Established The three interventionist arguments, for. Four, fulfilling human pioneer nature, held by Robert Zubrin. Rated weak: the premise is unsupported and the inference from nature to obligation is fallacious. Five, the species survival imperative, held by Joseph Gottlieb and by Elon Musk. Rated problematic, because cheaper terrestrial risk reduction dominates. Six, restoring Martian ecosystems, held by Christopher McKay. Rated the most plausible interventionist argument, because restoring indigenous microbial life evades both the scientific-conservation and the wilderness objection. Stoner's own verdict on the whole: “terraforming Mars is probably morally wrong.”

Frontier The adjacent positions, each citable and each with a holder. Sparrow characterises terraforming as “aesthetic insensitivity and the sin of hubris” and as “arrogant vandalism,” with a separate virtue-ethics treatment. Established That wording is quoted here at one remove, through Stoner and York: Sparrow's 1999 Environmental Ethics paper is paywalled and was not obtained for this brief, and no wording is attributed to the original text. Paul York advances a cosmocentric position — intrinsic value in inanimate nature, requiring a moral calculus between human and planetary interests — drawing on Keekok Lee on abiotic nature as a locus of intrinsic value. Christopher McKay holds a weak ecocentrism: an entire biosphere of alien life has far more value than individual microbes, and terraforming to support native life is defensible. Robert Zubrin holds pro-terraforming anthropocentrism, arguing that Martian microbes which never advanced never will, and may share a panspermia origin with terrestrial life and so not be uniquely valuable. Martyn Fogg offers a four-framework taxonomy — anthropocentrism, zoocentrism, ecocentrism, preservationism — and argues strict preservationism is untenable because human activity is itself a product of natural selection. Richard Sylvan holds a strong ecocentrism preserving native lifeforms on intrinsic-value grounds.

Frontier One empirical finding cuts into the debate from an unexpected direction. Wadsworth and Cockell's perchlorate result — irradiated soil killing Bacillus subtilis in thirty seconds — lowers forward-contamination risk, which weakens the scientific-conservation objection in its strongest form and therefore cuts for the interventionist side. It is reported here rather than buried in the soil-chemistry section because that is where it does its argumentative work.

Speculative And a point that the inventory result changes in the ethics. If global terraforming is closed by a fifty-fold shortfall, then much of the debate is about an act nobody can perform, and the live ethical question moves to the regional route — aerogel over a few square kilometres of ground ice within ±45 degrees latitude. That is a smaller act with the same structure, and none of the six arguments has been restated for it.

Established Terraforming is the largest thing the species has ever proposed to do, and the measurement that stops it is small and boring. The reservoir table fits on one page. A civilisational ambition of that size being closed by an accounting exercise is the most instructive thing about the subject.

Frontier The constraint is not technological, and that is what makes it decisive. Most limits in this corpus move with capability — better materials, cheaper launch, more power. This one does not: no technology creates carbon dioxide that is not on the planet. The correct civilisational reading is that Mars offers regional habitability and not a second Earth, and that no rate of progress changes it.

Frontier What survives is a staged approach that is already being pursued under a different name. Turyshev's conclusion — regional and enclosed habitability as the physically feasible route — is a description of Mars Colonization plus the aerogel line of research. The civilisational content of terraforming has migrated into habitat engineering, and the word has stayed behind.

Speculative The timescale spread is itself a civilisational fact and it explains public confusion. Warming is quoted at “a few decades” because it is a radiative-forcing problem; pressure at 10⁴ years because it is a mass-transport problem; loss at 10⁸ years because it is an escape problem; oxygen is unquantified because it is a biology problem nobody has run. Public claims about terraforming vary by seven orders of magnitude because different authors are answering different questions, and almost nobody says which.

Frontier And the ethical inheritance is unusually clear-eyed for a speculative field. A philosopher's survey rates the strongest argument on the board as the one against — avoiding colonialist vices, held by Sparrow and Lee — and rates both major arguments for as weak or problematic. Speculative A civilisation that decided to terraform Mars would be doing so against the balance of its own published moral reasoning, and the reasoning was done first.

These horizons separate four different rate-limiting processes, which is why published terraforming timescales differ by seven orders of magnitude:

• 10 yr: Frontier The tractable work is laboratory and modelling: a measured effective particle lifetime for engineered nanorods, an area-and-mass calculation for the aerogel route, and a chamber test of perchlorate remediation. All three are cheap and none is funded as a programme. Speculative A small aerogel field trial on a Mars lander is the only plausible flight milestone at this horizon and is not manifested. Established Nothing global is proposed, costed or scheduled by anybody.
• 25 yr: Frontier If the nanoparticle route were pursued, its own authors put ≥30 K of warming “within a few decades” — and the same paper puts the sustaining requirement at ~30 litres per second of engineered metal produced on Mars, indefinitely. Speculative Regional habitability under aerogel within ±45 degrees latitude is the defensible target at this range, and it is habitat engineering rather than terraforming. Handwave Any claim of a globally thickened atmosphere at this horizon contradicts a measured inventory shortfall of about fifty times.
• 50 yr: Speculative Pressure responds slowly even when warming does not: Ansari et al. give ~20% within months and then a further factor of 2–20 over a period “that could be as long as centuries.” Handwave The adsorbed-regolith reservoir cannot contribute on this horizon at all — thermal penetration through 100 m of regolith takes ~10⁴ years. Speculative Oxygen remains unquantified because no one has run the biology.
• 100 / 250+ yr: Handwave Beyond forecasting, and the honest structural statement is that the binding number is a planetary inventory rather than a rate of progress. Speculative At measured escape rates the present atmosphere would take ~2.5 × 10⁸ years to empty, so retention is the least urgent constraint on the list — and the most urgent, the ~50-fold CO₂ shortfall, does not move with time at all.

• Depends on One edge, and this brief keeps it while saying plainly that it is weaker than it looks. Artificial Magnetospheres owns loss; this brief owns inventory; and the sentence both pages print is that stopping atmospheric loss is not the same as having an atmosphere. The dependency is real as a narrative — a thick atmosphere would eventually need retaining — and on the measured numbers it does not bind: escape runs at 2–3 kg/s against ~2.3 × 10¹⁶ kg, about 250 million years to empty, and two measurement groups now hold that a magnetic field may not reduce escape at all, with Mars's ion escape characterised as supply-limited. What closes off the CO₂ route is Jakosky and Edwards' inventory, not the escape rate, and making that explicit is the most useful correction available to this page.
• Requires (not on this map) Three constraints that are not briefs on this map. The first is the only result that would reopen the global route: the reservoirs large enough to matter — deep carbonates at about 1 bar, clathrates under 150 mbar — are classified inaccessible, the barrier is thermodynamic rather than mechanical, and no mission has been proposed to test whether that classification is permanent. The second is the newest route's own stated major uncertainty, named by its authors: effective particle lifetime, on which the whole industrial requirement scales inversely and which is measurable in a chamber. The third is what that route actually asks for once the lifetime question is settled the unfavourable way — about 7 × 10⁵ m³ of engineered metal a year, described in the paper as a mine 700 m across lengthening 250 m annually, running on Mars indefinitely.
• Enables No enabling edge is claimed. Global terraforming enables nothing, because no surveyed abiotic pathway closes its eight constraints simultaneously and the CO₂ shortfall is a factor of about fifty. The part that survives — regional habitability under aerogel, measured at 45–50 K through 2–3 cm within ±45 degrees latitude — is already Mars Colonization's work under a different name, and Turyshev's own conclusion formally hands it across rather than claiming it here.
• Adjacent Artificial Satellites for Climate Control shares the toolkit at a different planet: one surveyed Martian warming route needs 10¹³–10¹⁴ m² of orbital reflector, two to twenty times the terrestrial sunshade whose mass arithmetic is worked out there and cited rather than repeated here. Space Law and Governance owns planetary protection and the question of who decides, while this brief owns the ethical arguments and their named holders. Climate Engineering is deliberately not adjacent: it owns Earth's climate and a different ethical literature, and the shared phrase “planetary engineering” is vocabulary rather than a seam.

Established “Terraforming Mars is a matter of engineering and will.” It is a matter of inventory. Everything mobilisable with present or near-future technology totals about 0.020 bar — polar caps 12 mbar, adsorbed regolith under 40 mbar, near-surface carbonates 12–150 mbar with limited access — against about 1 bar needed, and delivers less than 10 K against the roughly 60 K required. Frontier The reservoirs large enough to matter are the deep carbonates (~1 bar, inaccessible), the clathrates (inaccessible) and the more than 0.5 bar already lost to space (permanently gone). A better drill does not create CO₂ that is not there.

Established “That was one 2018 paper.” It was, and an independent 2026 systems analysis reproduces it from mass, forcing and throughput: a representative 20 mbar case yields 10 K or less, and “no surveyed abiotic global open-atmosphere pathway simultaneously closes pressure, composition, warming, power, throughput, retention, sink, and operations constraints.” Frontier Two independent analyses, eight years apart, different methods, same verdict.

Frontier “Nanoparticles solve it — they are five thousand times more efficient.” The >5000× figure is per unit mass in the atmosphere, comparing about 160 mg/m² of nanorods against about 1.1 kg/m² of an optimal greenhouse-gas mix. Established The number that belongs beside it is the paper's own operational requirement: particle fountaining at about 30 litres per second, roughly 7 × 10⁵ m³ of engineered metal a year, described by the authors as a mine of half-width 350 m lengthening 250 m annually. Speculative And there is an unresolved tension worth printing: the standing atmospheric inventory implied by 160 mg/m² over Mars's surface is about 23,000 tonnes, or ~860 m³ a year at the nominal 10-year particle lifetime — about 800 times less than the stated fountaining rate. The two reconcile only if effective residence time is days rather than years. Both inputs are the paper's; the ratio is derived here; the tension is stated rather than resolved.

Established “Nanoparticle warming is terraforming.” Its authors say twice, explicitly, that it is not: “Raising Mars’ temperature, by itself, is not sufficient to make the planet’s surface habitable for oxygenic photosynthetic life,” and “Nanoparticle warming, by itself, is not sufficient to make the planet’s surface habitable again.” Frontier They name the remaining barriers themselves — ~300 ppmw nitrates, very little O₂, and perchlorate-rich soil needing biological remediation — and they raise the risk that centuries of warming could drain ground water into empty pore space.

Handwave “Give Mars a magnetic field first and the rest follows.” Loss is the least urgent constraint on the list. Measured escape is 2–3 kg/s against an atmospheric mass of about 2.3 × 10¹⁶ kg — about 250 million years to empty. Frontier And whether a field would help is disputed: Gunell et al. find magnetisation increases total escape over a wide range of dipole strengths, and Ramstad and Barabash find Mars's ion escape supply-limited rather than driver-limited. This page keeps the dependency on Artificial Magnetospheres and states that it is not what binds.

Handwave “Nuke the poles” and “redirect comets.” Both circulate widely and neither appears as a costed proposal anywhere in the retrieved literature. Established The polar caps hold about 12 mbar — sublimating all of them completely still leaves the inventory short by a factor of roughly forty, which is the arithmetic that disposes of the first regardless of the method used.

Frontier “Water is the easy part — there is ice everywhere.” NASA's SWIM map covers buried ice from the equator to 60 degrees north within the upper metre, with a markedly non-uniform distribution, and gives no total inventory and no global equivalent layer. Established The 23-metre global-equivalent-layer figure that circulates is water lost since 4.2–3.5 Ga, not water present. No bulk inventory is asserted on this page because none was obtainable.

Established “Nitrogen is a detail.” A breathable buffer needs about 1.9 × 10¹⁸ kg of N₂; Martian soil carries about 300 ppmw nitrates. Frontier There is no polar cap of nitrogen and no identified reservoir of the required scale, and the problem receives a fraction of the attention the CO₂ problem gets.

Frontier “The ethics is just a set of vague worries.” It is six arguments with named holders and a published ranking. Preservationist: scientific conservation (strong), wilderness preservation after Rolston (strong), avoiding colonialist vices after Sparrow and Lee (compelling). Interventionist: pioneer nature after Zubrin (weak), species survival after Gottlieb and Musk (problematic), restoring Martian ecosystems after McKay (the most plausible of the three). Established Stoner's own verdict is that terraforming Mars is probably morally wrong — and Sparrow's exact wording is quoted on this page at one remove, through Stoner and York, because the 1999 original is paywalled and was not obtained.

Speculative “Mars will be terraformed within a human lifetime.” The only warming claim with a decadal timescale attached is nanoparticle warming, which its own authors say is not sufficient for habitability; pressure rise beyond ~20% is quoted as possibly centuries; the adsorbed reservoir takes ~10⁴ years to mobilise thermally; oxygen has no published timescale at all. Frontier The seven-order-of-magnitude spread in public timescale claims is not disagreement about Mars. It is different authors answering different questions without saying which.

Established And the framing itself. “Mars can be made habitable” is true regionally and false globally, and the split is measured rather than argued. Frontier Two independent analyses find no global abiotic pathway that closes; the same analyses identify regional and enclosed habitability as physically feasible; and the best laboratory result in the field is 2–3 cm of aerogel producing more than 45 K under Martian illumination. Speculative The honest position is that terraforming has already happened as a research programme — it turned into habitat engineering, and kept the older word for the parts that do not work.