A magnetic shield is a region of space, or a structure, that uses a magnetic field to divert charged particles away from something worth protecting. The phrase covers three quite different things at three scales: the Earth’s magnetosphere, the planetary-scale field that some engineers have proposed building at Mars to stop the solar wind stripping its atmosphere, and the mundane enclosures of mu-metal and copper that keep stray fields out of an MRI suite or a magnetometer. They share one piece of physics — a moving charge in a magnetic field feels a force perpendicular to both, so its path bends — and almost nothing else.[1][2]
The one we already have
Earth’s magnetosphere is generated by convective flow of molten iron in the outer core, coupled to the planet’s rotation. The resulting field is weak — a fraction of a gauss at the surface — but it extends a long way, and the solar wind, a supersonic plasma flowing at 300 to 800 km per second, sculpts it into a shape nothing like a bar magnet’s. On the sunward side the field is compressed to a boundary called the magnetopause, about ten Earth radii out, or roughly 63,000 km. A bow shock stands upstream of that, where the supersonic wind is slowed and heated. Downstream, the field is drawn out into a magnetotail that may extend a thousand Earth radii. Between the boundaries, charged particles are trapped on field lines in the Van Allen belts, bouncing from pole to pole.[3][4][5]
It is a shield, not a wall. Solar wind variations open temporary cracks and drive geomagnetic storms that can reach the ground and threaten power grids, and the coupling depends heavily on the orientation of the interplanetary field — when it points opposite to Earth’s, the impact of a coronal mass ejection grows. Most of the particles are deflected, a small fraction get through, and the atmosphere absorbs what the field misses. Life on Earth is protected by a system with holes in it, which is worth remembering whenever a proposed shield is described as complete.[6][7]
Mars as the counterfactual
Mars is what happens without one. It lost its global field around three billion years ago and retains only localised remnant magnetism in the southern hemisphere. Without a standoff, the solar wind reaches the ionosphere directly, and the MAVEN mission has measured the resulting escape — two main channels, one over the northern polar cap involving higher-energy ionospheric material and one equatorial, with oxygen ion escape reaching about 0.1 kg per second at times. Outgassing from the interior partly balances the loss, leaving a surface pressure of roughly 6 millibars. The planet is not airless, but it is a demonstration of what a missing magnetic shield costs over geological time.[8][9]
Building one for a planet
In 2017, NASA’s Jim Green and colleagues presented simulations of an artificial magnetic dipole parked at the Mars L1 Lagrange point, the balance point between the planet’s gravity and the Sun’s, about 320 Mars radii from the planet. A field large enough there would wrap Mars in its magnetotail, deflecting the solar wind and, in the simulations, cutting atmospheric loss by an order of magnitude during a hypothetical extreme coronal mass ejection. The follow-on climate modelling asked what happens if the loss stops: a rise of about 4 °C would be enough to sublimate the CO2 veneer at the north pole, thickening the atmosphere and beginning to melt trapped water. Green was careful about the word terraforming — nothing is added that Mars does not produce itself; the shield only lets a natural process run faster.[10][11]
The engineering literature is less lyrical. A 2021 analysis set out the physical and resource requirements and concluded that the lowest-power, lowest-mass option was not a coil at L1 at all but an artificial charged-particle ring around the planet — a manufactured radiation belt, possibly built by ejecting material from one of the moons, with electromagnetic and plasma waves driving a net current. It also noted a practical problem with the L1 design: magnetotails undulate with changing solar wind, so keeping a planet reliably inside a shield’s shelter is harder than a static diagram suggests. A 2025 perspective by Pelton and Green responded with three scenarios — a local field over a habitat, a system assembled from Phobos and Deimos, and a full L1 magnetosphere — and argued that such a project would need an international coalition, while noting that the same technology might one day be turned to protecting Earth from an extreme solar storm.[12][13]
A planetary magnetic shield is not a device you switch on. It is a piece of infrastructure the size of a moon’s orbit, and the honest papers on it read less like engineering plans than like feasibility studies for a civilisation’s budget.
The shield in the basement
At human scale the technology is mature and unglamorous. Passive shielding works by giving magnetic flux an easier path: a shell of high-permeability alloy — mu-metal is about 77% nickel and 15% iron, with relative permeability of 20,000 to 100,000 — collects field lines and carries them around the protected volume rather than through it. Because such alloys saturate, typically around 0.8 tesla, strong fields need a layered design with ordinary steel outside absorbing the bulk and mu-metal inside cleaning up the remainder. High-frequency fields are handled differently, by eddy currents in a conductive shell of copper or aluminium.[14][15]
Magnetoencephalography rooms show how far this is pushed. A commercial shielded room for optically pumped magnetometers uses four layers of mu-metal and one of copper, weighs more than ten tonnes, and still leaves a remnant internal field of 10–30 nanotesla, which demagnetisation coils bring down to 2–5 nT and active coils then cancel further. Superconducting shields go further, exploiting the Meissner effect to expel flux entirely and demonstrating shielding above one tesla — at the cost of cryogenics. MRI scanners use the active version: a secondary superconducting coil produces a field opposing the primary one, so a 2 T magnet plus a 0.5 T opposing coil gives a 1.5 T scanner with a much smaller stray field outside the room.[16][17][18]
Shielding a spacecraft
Between the planetary and the domestic sits the case that motivates most of the research: protecting crews from galactic cosmic rays and solar particle events on long missions. A magnetic field bends charged particles by the Lorentz force, and per unit mass high-temperature superconductors are the most efficient way to generate a useful field, with candidates including YBCO, BSCCO and magnesium diboride. Space is also a convenient place to run a superconductor, since the cold helps. Yet no crewed magnetic shield has flown, and a 2023 review of the field asks the obvious question — if this could remove one of the main barriers to interplanetary travel, why has it not been built? The answer is the usual combination: mass, cryogenics, and the difficulty of proving that a field which bends particles also reduces the dose that matters, without creating secondary radiation of its own.[19]
Select a year and month above to explore cold archived discussions.
Please select an archived topic above to view the discussion logs.