Artificial Magnetospheres

An artificial magnetosphere is a generated magnetic field placed around a spacecraft, a habitat or a planet to deflect charged particles — solar wind ions and, in the ambitious version, enough of the space environment to let a planet keep an atmosphere. The framing under test is that a planetary magnetic shield is buildable.

This brief has two findings, and the second is the more interesting one. The shield is not buildable at anything like the required scale, and the field's own engineering paper says so with its own numbers. But the literature has moved further than that: the premise that a planetary magnetic field protects an atmosphere is now actively disputed by the people who measure atmospheric escape, and the measured Martian loss rate implies a timescale of hundreds of millions of years, which makes the problem non-urgent even if the shield worked. The framing fails on the hardware and on the purpose, independently.

The page keeps two scales carefully apart. Spacecraft-scale magnetic shielding is a real research programme with costed hardware, laboratory validation and a sceptical independent review; its demonstrated protective distance is a few metres. Planet-scale shielding is a workshop abstract and a first-order sizing exercise. They share physics and nothing else, and the ratio between them is about a million. Where this brief ends and Terraforming begins is one sentence: this page owns loss, that page owns inventory, and stopping atmospheric loss is not the same as having an atmosphere.

Speculative The concept's primary source is a workshop abstract, and it is worth being precise about what it does and does not contain. Green, Hollingsworth, Brain, Airapetian, Glocer, Pulkkinen, Dong and Bamford presented A Future Mars Environment for Science and Exploration at the Planetary Science Vision 2050 Workshop in 2017 — authors from NASA Headquarters, Ames, Colorado, Goddard, Princeton and the Rutherford Appleton Laboratory. It proposes a dipole of “1 or 2 Tesla (or 10,000 to 20,000 Gauss)” from an inflatable structure at Mars L1, models surface-pressure scenarios of 10, 50, 100, 500 and 1,000 mbar against the present ~6 mbar, and states that beginning CO₂ polar-cap sublimation requires “an average change in the temperature of Mars of about 4 °C.” It describes it as “quite possible” that inflatable structures could generate the required fields.

Established What the abstract does not contain is the number everyone quotes from it. There is no standoff distance and no timescale for atmospheric thickening in it; the simulations “will be reviewed” and a projection “will be” made. The single most repeated claim about this concept — that Mars recovers a substantial atmosphere on a human-relevant timescale — is not in the source as a result. It is promised as future work. Any recovery timescale attributed to Green et al. is unsupported.

Frontier The engineering assessment came five years later, from the same community, and it is far more candid. Bamford, Kellett, Green, Dong, Airapetian and Bingham, How to create an artificial magnetosphere for Mars, Acta Astronautica 190 (2022). The requirement: 90–150 nT at the magnetopause for normal solar wind, 40–150 nT at Mars's orbital distance. That is a modest field. Everything that follows is about what it takes to make one at planetary scale.

Frontier Architecture A, a solid conductor loop. A solenoid on the Martian surface needs more than 5 gigaamperes to produce about 10⁻⁴ T at the surface. Moving it to areostationary orbit at six Mars radii cuts the current to under 0.2 GA for 460 nT. A carbon-nanotube solenoid of that kind draws about 10⁸ to 10¹¹ W and masses about 10⁷ kg — ten thousand tonnes — with a stored field energy of about 10¹⁷ J. Established The scale context is supplied by the authors themselves: global electricity consumption in 2020 was about 5.84 × 10¹⁷ J a year, and a DEMO-class fusion plant is about 2 GW around 2040. The upper end of the shield's draw is a substantial fraction of world electricity, delivered at Mars.

Frontier Architecture B, a plasma torus fed from a Martian moon. Minor radius about 2,000 km, major radius 2.8 Mars radii, a ring current of about 7 × 10⁷ A at a current density around 5 × 10⁻⁵ A/m², sustained by about 15 kg per orbit per loop of ion supply — with Phobos material lasting on the order of 10¹¹ Earth years. Speculative The authors state that the torus faces unsolved diffusion and solar-wind pickup problems and needs much further research.

Handwave The core-restart alternative is priced in the same paper and dismissed in it. Re-melting Mars's core would take energy equivalent to about 10¹¹ one-megaton hydrogen bombs, and the authors state it is “unlikely that this option would be feasible.”

Frontier The most defensible part of that paper is its list of its own caveats, and they should be read as the authors' verdict. They record that “the volume involved is vast” for a planetary-scale solid conductor; that Mars's “very uneven gravitational pull” means an orbiting ring “would require considerable delta-V to maintain the orbit”; that Mars L1 is a “very shallow gravitational ‘island’” whose position drifts “by a considerable distance during the 687 days Martian year” because of orbital eccentricity; that “resources needed would be vast”; and that the calculations are made “to first order … as higher precision figures would be meaningless without comparable precision for engineering.” Established This is the field's only engineering study, and it is a first-order sizing exercise that its own authors decline to present as a design. Interest runs against the finding here — proponents publishing the numbers that defeat their own concept — which is why it is weighted heavily.

Frontier Now the second finding, which is the one the popular framing has not absorbed: a magnetosphere may not help. Gunell, Maggiolo, Nilsson and colleagues, in Astronomy & Astrophysics in 2018, modelled seven escape mechanisms as a function of dipole moment and found that magnetisation increases total escape over a wide range of dipole strengths, because a magnetosphere opens polar-cap and cusp escape channels that an unmagnetised planet does not have. Escape peaks “at the transition between magnetised and unmagnetised planets,” with polar-cap escape dominating for hydrogen and cusp escape for oxygen. Their title states the conclusion: an intrinsic magnetic field does not protect a planet against atmospheric escape.

Frontier Their measured comparison is the part that does the damage. Observed mass escape rates: Venus, unmagnetised, about 0.5 kg/s; Earth, magnetised, about 1.4 kg/s; Mars, unmagnetised, 0.7–2.1 kg/s. In their words, the rates “from these three planets are similar (in the approximate (0.5–2) kg s⁻¹ range),” and “magnetisation is not a sufficient condition for protecting a planet from atmospheric loss.” Established Earth, with a magnetosphere, loses mass roughly three times faster than Venus without one. That single comparison does more damage to the framing than any cost estimate.

Frontier The review-level version is sharper still, and it names the mechanism. Ramstad and Barabash, in Space Science Reviews in 2021, work from ion-flux measurements at all three planets and conclude, verbatim, that “an intrinsic magnetic dipole field is not required to prevent stellar wind-driven escape of planetary atmospheres, and the presence of one may instead increase the rate of ion escape.” Their framework is that escape is limited by bottlenecks — ion supply, solar-wind energy transfer, transport efficiency — and which one binds differs by planet. Venus and Earth are energy-limited; Mars is supply-limited because of its low gravity, and ion escape therefore contributed relatively little to early atmospheric loss compared with neutral escape.

Frontier That is the deepest cut available against the concept, and it is worth stating slowly. If Mars's ion escape is supply-limited, then reducing the solar-wind driver — which is exactly what an L1 dipole does — does not reduce the loss proportionally, because the driver is not the binding constraint. The shield addresses a bottleneck that is not the bottleneck.

Established And the loss it would prevent is slow. This is the arithmetic the popular coverage never does. MAVEN's measured present-day rates, from Jakosky et al. in Icarus (2018): total combined hydrogen and oxygen loss about 2–3 kg/s; hydrogen Jeans escape 160–1,800 g/s and strongly seasonal; photochemical hot-oxygen escape about 1,300 g O/s; oxygen ion loss about 130 g O/s; sputtering about 80 g O/s. Integrated loss from 4.2–3.5 Ga exceeds 0.8 bar of CO₂ or a 23 m global equivalent layer of water, and argon isotopes independently require that about 70% of all argon ever in the atmosphere has been lost. The 2015 NASA announcement figure, still the most quoted, is about 100 grams per second of ion escape, rising significantly during solar storms, distributed roughly 75% down the tail and 25% in the polar plume — with Bruce Jakosky's framing that it is “like the theft of a few coins from a cash register every day.”

Established Now the division, derived here with both inputs given. Mars's atmospheric mass from published constants — surface pressure 600 Pa, gravity 3.71 m/s², surface area 1.448 × 10¹⁴ m² — is pressure times area divided by gravity, or about 2.3 × 10¹⁶ kg. At the measured total escape rate of 3 kg/s, emptying the present atmosphere takes about 7.8 × 10¹⁵ seconds, or roughly 250 million years. At the 100 g/s ion-escape figure alone it is about 7 billion years. Established The loss timescale is 10⁸ to 10⁹ years. The derivation is this brief's; the atmospheric mass is computed rather than quoted, because the NASA fact sheet was not obtainable during research. Atmospheric stripping does not act on the timescale of a settlement, a terraforming project, a civilisation or a species. A shield built to stop it would be protecting a tank that takes a quarter of a billion years to drain.

Established It also does not solve the radiation problem, which is what most readers assume it solves. MSL/RAD measurements during cruise to Mars give 1.8 mSv per day and 0.66 Sv for a round-trip transit, of which about 95% is galactic cosmic rays and only about 5% solar energetic particles. Cary Zeitlin's own line: “even an aluminum hull a foot thick wouldn’t change the dose very much.” Frontier The inference, drawn carefully here rather than quoted from that paper: galactic cosmic ray primaries reach many GeV, and both bulk shielding and magnetic deflection scale badly against them; what actually attenuates cosmic-ray dose at a planetary surface is atmospheric column mass, not field geometry. An L1 dipole would address atmospheric stripping (slow) and solar energetic particles (5% of dose), and leave the dominant hazard largely untouched.

Frontier Does an intrinsic field protect an atmosphere at all? This is now the field's live question and it was a textbook assumption five years ago. Gunell et al. find magnetisation increases net escape over a wide range of dipole moments; Ramstad and Barabash find a dipole is not required and may increase ion escape; the measured rates at three planets sit within a factor of four of each other regardless of magnetisation. The whole concept's purpose is under dispute in the peer-reviewed literature, and the dispute is between measurement groups rather than between modellers.

Frontier Which bottleneck binds at Mars? The supply-limited versus energy-limited distinction is the most consequential open question here. If Mars's ion escape is supply-limited, shielding the planet from the solar wind reduces loss by much less than proportionally. Nothing in the retrieved literature quantifies what an artificial shield would do to a supply-limited escape channel, and until somebody does, the concept has no stated benefit at all.

Frontier The unshielded case is still being discovered. Zhang, Barabash, Holmström and colleagues showed in Nature in 2024 that when the solar-wind cone angle is small — around 4 degrees — Mars's induced magnetosphere degenerates: no dayside shock, only weak flank shocks, a cross-flow plume, ambipolar-field-driven upstream ion flow, and solar-wind particles reaching the ionosphere directly, with hybrid simulations matching MAVEN and Mars Express observations. Their own closing caveat: “It remains to be studied what the secondary effects are on processes like atmospheric loss through ion escape.” Established The plasma physics of the case you would be intervening in was still producing surprises in 2024, which is a poor foundation for an engineering programme.

Frontier How far does a mini-magnetosphere actually reach? Bamford and colleagues' 2014 laboratory work argues that the surrounding plasma does most of the work, cutting the power that vacuum calculations imply: surface fields of order hundreds of nanotesla suffice, simulations give 100% exclusion for particles at 10⁵ times background plasma energy and 95% at 10⁶ times, and the mass-loading requirement is strikingly small — about 131 g of xenon per storm event, under half a kilogram for three solar particle events across an 18-month mission. Frontier These are proponents on their own concept, and the scope matters: it is solar energetic particles, the 5% of transit dose, not cosmic rays.

Established And the scale that has actually been demonstrated is small enough to state exactly. Silva et al.'s bench-validated hybrid simulations give plasma density 10¹² cm⁻³ at 5 eV, flow 400–620 km/s, fields of 0.01–0.4 T, and a magnetopause standoff of 26.7 mm predicted against 28.5 mm measured at 0.2 T, with incoming plasma totally deflected at the standoff. Scaled to deflect 1 MeV protons in space they estimate 0.72 T from a current loop with a magnetic moment of about 7.2 × 10⁶ A·m², giving a protective distance of “a few meters.” That is the demonstrated and modelled scale of this technology. The planetary concept asks for a standoff of order a planetary radius — the same physics at a ratio of about 10⁶.

Frontier Is active shielding better than passive mass? This is the field's motivating claim and it is not established. NASA's NIAC Phase II study explicitly did not perform the comparison. The independent 2023 review by Ferrone and colleagues finds confined magnetic fields infeasible because their mass exceeds comparable passive shielding, dismisses electrostatic shields on secondary-particle grounds, rejects plasma shields because radiation-belt degradation reduces effectiveness, and leaves unconfined fields with deployed superconducting magnets as the only promising family — at a stored energy of about 10¹⁵ J.

Speculative What do the natural experiments say? Mars retains strong remanent crustal magnetisation concentrated in Terra Sirenum and Terra Cimmeria, described as much stronger than anywhere else on Mars or in terrestrial crustal fields measured from orbit, with about 2,000 nT measured at the surface at the InSight site and orbital models systematically under-predicting surface values — implying unresolved small-scale magnetisation. The dynamo ceased by roughly 3.6–4.1 Ga. Established But the review that establishes those field strengths does not quantify mini-magnetosphere extent or local atmospheric protection, so no claim about crustal fields protecting local atmosphere is made on this page.

Speculative And the question that would reframe everything if answered: whether the terraforming community still regards a shield as a precondition at all. The field's 2025 research-agenda paper sets out a warming-first roadmap — warming, then engineered oxygenic photosynthesis, then oxygen build-up — and does not name a magnetic shield among its preconditions or research priorities in the abstract obtained here. Frontier That is an absence in an abstract rather than a stated rejection, and it should be checked against the full text; but if it holds, it is evidence about how the community now weights the loss problem.

Frontier Mass is the first bottleneck and it is not close. Bamford's carbon-nanotube solenoid masses about 10⁷ kg — roughly a thousand times the total mass of everything humans currently have in Earth orbit, on ESA's measured baseline of 10,000–12,000 tonnes. Nothing about that is an electromagnetism problem.

Frontier Power is the second. 10⁸ to 10¹¹ W, continuously, at Mars. The upper end is a substantial fraction of world electricity consumption; the lower end is fifty DEMO-class fusion plants. The paper gives the range across three orders of magnitude because, at first-order precision, that is the honest range.

Handwave Current is the third, and the surface architecture fails on it outright. A ground-based solenoid needs more than 5 gigaamperes. Moving to areostationary orbit cuts it to under 0.2 GA, which is progress of a kind — but 0.2 GA is still four orders of magnitude beyond any current ever carried in a superconducting magnet.

Frontier Station-keeping is a modest delta-v with an immodest attachment. Sun–Mars L1 and L2 sit about one million kilometres from Mars, and libration orbits there are described as “precarious” requiring “precise and continuous stationkeeping” at about 2 m/s per year — against roughly 50 m/s per year for low Mars orbit and 200 m/s per year for areosynchronous. Established The delta-v is small; the object is 10⁷ kg, it must be paid forever, and Mars's orbital eccentricity drifts the point by a considerable distance across the 687-day year.

Frontier The scale gap between demonstration and requirement is the bottleneck nobody can engineer around. Demonstrated protective distance: a few metres. Required standoff for a planetary shield: of order a planetary radius. That is a factor of about 10⁶, and the intervening regime — tens of metres to thousands of kilometres — has no experiment in it.

Established And the bottleneck that undercuts the others: no stated benefit. The escape timescale is 10⁸ to 10⁹ years, the physics of whether a field helps is disputed, Mars's ion escape may be supply-limited rather than driver-limited, and 95% of transit radiation dose is cosmic rays against which magnetic deflection scales badly. A project with a 10⁷ kg mass requirement and no quantified benefit is not bottlenecked on engineering.

Established The single adjudicated dependency is fabrication, and it is the right one. A 10,000-tonne carbon-nanotube solenoid drawing 10⁸–10¹¹ W is a manufacturing and assembly problem at planetary scale, which is Space-Based Manufacturing's territory. That brief's demonstrated scale is laboratory instruments hardened for a station, not factories, and the largest orbital-assembly job anyone has costed there is 5.9 million kilograms — less than the mass of one Mars shield.

Frontier The scientific dependency is a quantified answer to the escape question. Two measurement groups now hold that magnetisation does not protect and may increase ion escape. Until somebody computes what an artificial dipole does to a supply-limited escape channel at Mars, the concept has no benefit figure to put against its cost figure, and no engineering programme can be specified without one.

Established And the monitoring dependency has just lapsed. MAVEN — the source of every loss-rate number on this page — lost signal on 6 December 2025 and the mission formally ended on 3 June 2026, after twelve and a half years. The measurements stand; the time series has stopped. Any refinement of the escape rates, or any test of the Zhang et al. degenerate-magnetosphere result against new solar-wind conditions, now waits on a mission that does not exist.

Frontier Superconductor performance is the enabling materials dependency for the only architecture the independent review does not reject. Ferrone et al. leave unconfined fields with deployed superconducting magnets as the sole promising family and put the stored energy for a fully deployed shield at about 10¹⁵ J. That is a materials-and-cryogenics dependency developed elsewhere for other reasons.

Established The reciprocal dependency runs to Terraforming, and it needs stating honestly rather than assumed. That brief formally depends on this one, and this pack weakens the edge: on the measured numbers, loss is not what binds terraforming. The constraint that closes off the Martian CO₂ route is inventory — about 20 mbar mobilisable against about 1 bar needed. The shield is a retention precondition for a hypothetical thick atmosphere on geological timescales, not a near-term one.

Established The experiment that has been done is a bench-scale plasma deflection, and it worked. Silva et al.'s hybrid simulations predicted a 26.7 mm magnetopause standoff against 28.5 mm measured at 0.2 T, with incoming plasma totally deflected at the standoff distance. The physics is validated. The scale is millimetres.

Frontier The costed hardware experiment is spacecraft-scale and it exists on paper. NASA's NIAC Phase II MAARSS study gives an 8 T·m baseline at 1 T over an 8 m coil diameter, about 3,240 kg of coil structures, 410 MJ stored and 43.5 kA; and a 25 T·m configuration at 1.5 T over 16.7 m, about 16,837 kg of strongback and 5,440 MJ. Dose reduction for the 20 T·m configuration is about 65–69% against baseline. Frontier The study's own stated limit is that a detailed comparison with passive shielding was not performed in that phase — which is the comparison the whole concept rests on. Treat the paradigm-shift language as NIAC advocacy and the masses as engineering estimates.

Frontier The decisive missing experiment is an orbital mini-magnetosphere. Nothing has flown. A deployed superconducting loop generating a measurable standoff against the real solar wind, with an instrumented particle count inside and outside, would move the whole field from simulation to measurement and would test the plasma-assistance argument that makes the power budget tractable.

Frontier The escape-physics experiment is a computation, not a spacecraft. What an artificial dipole of a given moment does to each of the seven modelled escape channels at Mars — including the polar-cap and cusp channels that magnetisation opens — is calculable with the codes Gunell et al. already used. It has not been published, and it would either give the concept a benefit figure or remove its purpose.

Speculative The natural experiment is on the surface already. Mars's crustal magnetisation reaches about 2,000 nT at the InSight site and orbital models under-predict surface values, implying unresolved small-scale structure. A measurement of ionospheric and neutral density above a strong crustal anomaly versus a null region would be a direct empirical test of whether a local field protects local atmosphere — and the retrieved crustal-field review does not report one.

Established What cannot be experimented on is the planetary case. There is no sub-scale version of a planetary magnetosphere: the standoff distance scales with the field and the solar-wind pressure, and a few metres and a few thousand kilometres are not the same regime. The field's evidence will remain laboratory-plus-simulation until somebody flies something, and nobody has proposed to.

Frontier Requirement one: 90–150 nT at the magnetopause. This is the honest, modest number at the centre of the subject, and it is the reason the concept keeps being taken seriously. Generating it is trivial in a laboratory and requires 10⁷ kg at Mars, and the difference is entirely geometry: field falls as the cube of distance and the required standoff is a planetary radius.

Handwave Requirement two: current. Above 5 GA for a surface solenoid, under 0.2 GA at areostationary orbit. The largest currents ever carried in superconducting magnets are in the tens of kiloamperes. The areostationary option is four orders of magnitude beyond that and is the better of the two.

Frontier Requirement three: continuous power at 10⁸–10¹¹ W, delivered at Mars, indefinitely. The stored field energy alone is about 10¹⁷ J. Any quench of a magnet holding 10¹⁷ J is an energy-release event with no engineering precedent, and no retrieved source discusses fault modes.

Frontier Requirement four: hold position for the design life. About 2 m/s per year at Sun–Mars L1 — small — applied to 10⁷ kg, forever, at a point one million kilometres from the planet whose position drifts considerably across a 687-day year because of orbital eccentricity. The delta-v is the easy part and the mass makes it expensive anyway.

Frontier The spacecraft-scale requirements are real and much smaller. MAARSS's 8 T·m baseline is 3,240 kg of coil structure at 43.5 kA and 410 MJ for roughly two-thirds dose reduction. That is a hard but recognisable aerospace problem — and the independent review's verdict is that even this family is not yet ready for development funding.

Frontier And the requirement the mini-magnetosphere line of work makes cheap, if it holds. Bamford et al. argue the surrounding plasma does most of the deflection, so surface fields of hundreds of nanotesla suffice and mass loading is about 131 g of xenon per storm event, under half a kilogram for a whole mission. Speculative If that scales even to a habitat, it is the most consequential result in the subject. It has been demonstrated at a few metres by its own proponents.

Established Space-Based Manufacturing is the single adjudicated edge and the seam is clean. This brief owns the plasma physics and the field-strength requirement; that brief owns whether a 10,000-tonne carbon-nanotube solenoid can be built anywhere at all. Its demonstrated scale is laboratory instruments hardened for a station, and the largest orbital-assembly job anyone has costed there is less than the mass of one shield.

Established Terraforming depends on this brief, and the dependency needs its weakness stated. This page owns loss; that page owns inventory. The sentence both need: stopping atmospheric loss is not the same as having an atmosphere. Frontier And on the measured numbers the edge is narrative rather than binding: loss runs at 2–3 kg/s against 2.3 × 10¹⁶ kg, and the constraint that actually closes off the Martian CO₂ route is the inventory measurement, not the escape rate.

Frontier Space Weather Engineering defers the planet-scale case to this page, and that pointer is correct. The near-Earth active-shielding question — spacecraft, habitats, storm shelters — sits between the two, and the useful division is that this brief owns the physics of magnetic deflection and its scale limits wherever they appear.

Established No edge to Climate Engineering, and the rewrite should resist one. That brief owns Earth's energy budget and the politics of deliberate intervention; this is a plasma-physics and megastructure question about another planet with no terrestrial governance content. If the two look adjacent it is because both are called “planetary engineering,” which is a word rather than a seam.

Frontier Superconducting magnet technology is the shared substrate. Every non-rejected architecture in the independent review uses deployed superconducting magnets, and the fusion programme is developing them at scale for unrelated reasons — which means the most likely route to a flyable magnetic shield runs through Commercial Fusion's magnet supply chain rather than through anything in this brief.

Established Planetary protection and the authority to re-engineer another planet's environment belong to Space Law and Governance and to Terraforming. One clause here, and no more.

Established The institutional record of this concept is one workshop abstract, one journal paper, and no programme. The 2017 abstract came from a Planetary Science Vision 2050 workshop — a forum for long-horizon ideas, not a funding decision — and the 2022 Acta Astronautica paper is a first-order sizing exercise by the same community. No agency has a line item for a planetary magnetic shield.

Frontier NIAC is the institutional home of the spacecraft-scale version and it funds concepts. MAARSS reached Phase II in 2014 with credible masses and fields and an explicit statement that the comparison with passive shielding was not performed. An advocacy-framed study that declines the decisive comparison is a characteristic NIAC output, and it is why the independent review matters more.

Frontier The independent verdict is that even the small version is not fundable yet. Ferrone, Willis, Guan, Ma, Peterson and Kry — a radiation-physics group with no stake in the concept — conclude after sixty-two years of the idea that magnetic shielding “has remained primarily in conceptual stages,” and that substantial simulation, prototyping and sensitivity analysis are needed before space agencies should commit to development funding. Interest neutral, verdict negative.

Established Interest disclosure is unusually easy to apply here, and it changes how the evidence reads. The concept's proponents published the 10⁷ kg, the 5 GA and the 10¹¹-megaton core restart themselves; the mini-magnetosphere results come from the same proponents and are favourable to them; the escape-physics results come from measurement groups with no stake either way. The pattern is that everything favourable to the concept comes from inside it and everything unfavourable comes from both inside and outside.

Established The most consequential institutional event in this subject during the last year was a spacecraft going quiet. MAVEN lost signal on 6 December 2025 and the mission ended on 3 June 2026. Every measured escape rate on this page comes from it, and there is no successor monitoring the Martian escape channels.

Speculative What an institution would need to move this from abstract to programme. A quantified benefit — what a dipole of stated moment does to measured escape at Mars — and a flown mini-magnetosphere at spacecraft scale. The first costs a graduate student and a code; the second costs a small mission. Neither has been commissioned, and the absence of the cheap one is more informative than the absence of the expensive one.

Frontier The ethical weight of this subject sits almost entirely on the spacecraft-scale case, where the beneficiary is a crew. Transit dose is 0.66 Sv for a round trip at 1.8 mSv/day, and a technology that removes even the 5% solar-particle share removes the acute risk while leaving the chronic one. Radiation protection for crews is a genuine duty and the honest reading is that magnetic shielding is not yet the way to discharge it.

Speculative The planetary case raises the question of who authorises re-engineering another planet's environment, and this brief does not own it. Deliberate modification of Mars's plasma environment is a planetary-protection and authority question belonging to Space Law and Governance and to Terraforming, where the ethical arguments have named holders. One clause here, and a pointer.

Frontier There is a communication ethics problem specific to this concept and it is measurable. The claim that Mars would regain a substantial atmosphere on a human-relevant timescale is attributed to the 2017 abstract in wide circulation, and that abstract contains no such projection — it promises one as future work. Established A number that does not exist in its cited source has propagated for nine years, and the only remedy available to a page like this one is to say so explicitly.

Frontier The opportunity-cost argument is unusually strong here because the alternative is cheap. Passive shielding, shorter transits and storm shelters are available now; the independent review finds confined-field architectures heavier than comparable passive shielding. Spending on an unproven active system rather than on the demonstrated passive one is a real trade-off with a real cost in delivered protection.

Speculative And a scientific-conservation point that belongs to this page rather than to the ethics literature next door. Mars's present plasma environment is a natural laboratory for how unmagnetised planets lose atmospheres — a question that bears on habitability everywhere. Placing an artificial dipole at Mars L1 would destroy the experiment while testing a hypothesis about it, which is a specific and unusual form of the conservation objection.

Established The civilisational claim attached to this concept is that a shield is the first step to making Mars habitable, and the measured numbers dissolve it. Loss runs at 2–3 kg/s against an atmosphere of 2.3 × 10¹⁶ kg — about 250 million years to empty. No civilisation needs to solve this problem in order to do anything else.

Frontier What the concept actually measures is industrial capacity, not planetary science. Ten thousand tonnes at Mars is about a thousand times everything in Earth orbit; 10⁸–10¹¹ W is between fifty fusion plants and a substantial fraction of world electricity. A civilisation able to build it would already have solved the problems that make Mars hard.

Speculative The spacecraft-scale version has a genuine civilisational reading and it is modest. If plasma-assisted mini-magnetospheres work at habitat scale for a few hundred grams of xenon per storm, they make long-duration crewed operations meaningfully safer against solar events. That is a real capability at a real scale, and it is not the thing the phrase “artificial magnetosphere” conjures.

Frontier The most durable civilisational lesson here is epistemic. A concept was proposed in a workshop abstract, acquired a projection that the abstract never made, and circulated for nine years; its own proponents then published an engineering assessment that defeats it, and two independent measurement groups undermined its purpose. The correction came from inside the field and from measurement, and it took less than a decade. That is the system working.

Speculative And the reframing that survives. If magnetisation does not protect an atmosphere — if Earth loses mass faster than Venus and Mars's escape is supply-limited — then the standard story about why Mars is dry needs revising, and so does the habitability filter applied to exoplanets around active stars. The interesting consequence of this literature is not about engineering Mars at all; it is about what a magnetic field is for.

These horizons track measurements and reviews, because there is no shield programme anywhere to track:

• 10 yr: Frontier The escape-physics dispute is the thing that will actually move: whether magnetisation increases net escape, and whether Mars's ion escape is supply-limited, are calculable questions with active groups on them. Established MAVEN ended on 3 June 2026, so refinement of the loss rates now waits on a mission nobody has proposed. Speculative A flown spacecraft-scale mini-magnetosphere is possible at this horizon and is not funded; the independent review says the family is not ready for development funding. Handwave No planetary shield element is scheduled, proposed as a mission, or costed beyond first order.
• 25 yr: Frontier If deployed superconducting magnets mature through fusion and other programmes, an unconfined-field spacecraft shield becomes a plausible flight demonstration at this range. Speculative A habitat-scale storm shelter using plasma-assisted deflection is the most optimistic defensible milestone, and it protects against the ~5% solar-particle share of dose rather than the ~95% cosmic-ray share. Handwave Nothing at planetary scale has a milestone at this range.
• 50 yr: Handwave No basis for a planetary forecast. The requirement is 10⁷ kg and 10⁸–10¹¹ W at another planet, and the only route to it runs through an orbital manufacturing capacity whose largest costed job is smaller than one shield. Speculative The more likely development is that the question is dropped rather than solved, because the escape-physics literature removes its purpose.
• 100 / 250+ yr: Handwave Beyond forecasting, and the timescale comparison is worth stating literally: at measured rates Mars's atmosphere empties in about 250 million years, so a shield deferred by a century costs about four hundred-thousandths of the reservoir it protects. Speculative The defensible structural statement is that this is one of the few engineering subjects where waiting is nearly free.

• Depends on One edge, and it is the right one because nothing here is blocked on physics. The field requirement is 90–150 nT at the magnetopause, which is modest; what defeats the concept is that producing it at a planetary standoff needs about 10⁷ kg of carbon-nanotube solenoid drawing 10⁸–10¹¹ W, on its own proponents' figures. That is a fabrication-and-assembly problem, and Space-Based Manufacturing owns it. The scale comparison is the seam: a Mars shield is about a thousand times the total mass humans currently have in Earth orbit, and heavier than the largest orbital-assembly job anyone has costed anywhere.
• Requires (not on this map) Three constraints that are not briefs on this map, and the last two are why the first has never been attempted. The industrial constraint is the 10⁷ kg structure itself. The first scientific constraint is a missing benefit figure: Gunell et al. find magnetisation increases net escape across a wide range of dipole moments, and Ramstad and Barabash find Mars's ion escape supply-limited rather than driver-limited — so nobody has computed what an artificial dipole would actually do, and the concept has no quantified benefit to weigh against its cost. The second is a missing measurement: the demonstrated and modelled protective distance of this technology is “a few meters,” bench-validated at a 26.7 mm predicted against 28.5 mm measured standoff, and the planetary concept asks for a standoff of order a planetary radius — the same physics at a ratio of about 10⁶, with no experiment anywhere in between.
• Enables Terraforming records a dependency on this brief, and this page states the edge's weakness rather than quietly keeping it: stopping atmospheric loss is not the same as having an atmosphere, measured escape runs at 2–3 kg/s against 2.3 × 10¹⁶ kg for a ~250-million-year emptying time, and two measurement groups now hold that a magnetic field may not reduce escape at all. The retention argument is real on geological timescales and is not what binds terraforming on any human one.
• Adjacent Space Weather Engineering defers the planet-scale case here and keeps the near-Earth one, which is a correct division. Commercial Fusion is adjacent through hardware rather than purpose: every architecture the independent review does not reject uses deployed superconducting magnets, and the supply chain for those is being built for fusion. Mars Colonization owns the sealed-habitat alternative, which is what the radiation numbers on this page actually point at.

Established “NASA proposed a magnetic shield that would let Mars rebuild its atmosphere in a few centuries.” The 2017 workshop abstract contains no standoff distance and no recovery timescale. It says the simulations “will be reviewed” and a projection “will be” made. Frontier The most repeated claim about this concept is not in its own primary source as a result, and any recovery timescale attributed to Green et al. is unsupported by the document.

Established “Give Mars a magnetic field and it keeps its air.” Twice unsupported. First, the air is leaving on a 10⁸-year clock: 2–3 kg/s measured against an atmospheric mass of about 2.3 × 10¹⁶ kg is roughly 250 million years to empty, derived here from published surface pressure, gravity and area. Frontier Second, a field may not slow it. Gunell et al. find magnetisation increases total escape across a wide range of dipole strengths by opening polar-cap and cusp channels, and report Earth at ~1.4 kg/s magnetised against Venus at ~0.5 kg/s unmagnetised.

Frontier “An artificial magnetosphere would protect astronauts from radiation.” About 95% of transit dose is galactic cosmic rays — 1.8 mSv/day, 0.66 Sv round trip — against which both magnetic deflection and bulk shielding scale badly; Zeitlin's own line is that a foot of aluminium would not change the dose much. Speculative A shield addresses the ~5% solar-particle share. The proponents' own mini-magnetosphere work is explicitly scoped to solar energetic particles, and this page keeps that scope.

Handwave “We could just restart the Martian core.” Priced in the same paper that proposes the shield: energy equivalent to about 10¹¹ one-megaton hydrogen bombs, with the authors' own conclusion that it is “unlikely that this option would be feasible.” Established The dynamo ceased by roughly 3.6–4.1 Ga and nothing in the retrieved literature proposes a mechanism for restarting it.

Speculative “The L1 shield is an engineering programme.” It is a first-order sizing exercise whose authors state that “higher precision figures would be meaningless without comparable precision for engineering,” that “the volume involved is vast,” that “resources needed would be vast,” and that the plasma-torus option needs much further research. Frontier They also give the numbers — 5 GA, 10⁷ kg, 10⁸–10¹¹ W — which is what makes the paper valuable and what makes the concept fail.

Established “Mini-magnetospheres have been demonstrated, so the planetary case is a scaling problem.” The demonstrated and modelled protective distance is “a few meters,” from bench work that matched a 26.7 mm predicted standoff against 28.5 mm measured. Frontier Green's concept asks for a standoff of order a planetary radius. That is the same physics at a ratio of about 10⁶, and there is no experiment anywhere in between.

Speculative “Active shielding beats passive mass shielding per kilogram.” This is the field's motivating claim and it is not established. NASA's NIAC Phase II study explicitly did not perform the comparison, and the interest-neutral 2023 review finds confined-field architectures infeasible because their mass exceeds comparable passive shielding, with the whole family “primarily in conceptual stages” after sixty-two years.

Frontier “Turning down the solar wind reduces the loss proportionally.” Not if the escape is supply-limited. Ramstad and Barabash characterise Venus and Earth as energy-limited and Mars as supply-limited because of its low gravity, meaning the solar-wind driver is not the binding constraint at Mars. Speculative An L1 dipole addresses a bottleneck that may not be the bottleneck, and no published calculation quantifies what it would do instead.

Speculative “Mars's crustal fields already show local shielding works.” Mars does retain strong remanent magnetisation in Terra Sirenum and Terra Cimmeria, with about 2,000 nT measured at the surface at the InSight site and orbital models under-predicting surface values. Established But the review that establishes those strengths does not quantify mini-magnetosphere extent or local atmospheric protection, so no such claim is made here.

Frontier “We understand the unshielded case well enough to intervene in it.” In 2024 a Nature paper showed that at small solar-wind cone angles Mars's induced magnetosphere degenerates entirely — no dayside shock, solar-wind particles reaching the ionosphere directly — and its authors close by saying the consequences for ion escape remain to be studied. Established The baseline is still being characterised.

Frontier And the framing itself. “A planetary magnetic shield is buildable” fails twice and the two failures compound. Established The thing is not buildable at 10⁷ kg and 10¹¹ W, on figures published by the people who want to build it. Frontier And the problem it would solve is neither urgent — 250 million years — nor clearly solved by a magnetic field, on measurements from groups with no stake in the answer. Speculative The most interesting thing in this subject is no longer whether we could build a magnetosphere. It is that the field is no longer sure what one is for.