Pulsar Magnetosphere

A magnet that refuses to sit in a vacuum

Take a neutron star — a city-sized ball of nuclear matter about 12 km across and 1.4 times the Sun's mass — and spin it once a second while burying a magnetic field of 10¹² gauss inside it, a trillion times the field of a refrigerator magnet. Now ask the obvious question: what is the space around it like? The intuitive answer, a clean vacuum dipole field rotating in empty space, is wrong, and the reason it is wrong is the most important fact in this whole subject.

A rotating magnet generates an electric field. For a conductor — and a neutron star's crust is an excellent conductor — that induced field has a component along the magnetic field at the surface, and its strength is staggering: the potential drop across a polar cap is of order 10¹⁶ volts. No surface binding energy can hold a charge against a field that strong. The star tears electrons and ions off its own surface and flings them into the surroundings until the space fills with plasma. This is the central insight of Peter Goldreich and William Julian's 1969 paper: a pulsar magnetosphere is not a vacuum. It is a plasma-filled, co-rotating, current-carrying structure, and everything a pulsar does — the radio beam, the gamma rays, the slow spin-down, the surrounding nebula — happens inside it.

How it works: co-rotation, the light cylinder, and open field lines

Once plasma fills the magnetosphere, it is locked to the field lines and forced to co-rotate with the star — the field is so strong that the plasma has essentially no choice. Co-rotation is fine close in, but it runs into a hard limit. A point co-rotating at distance r from the spin axis moves at speed v = Ω r, where Ω = 2π/P is the angular spin rate. That speed reaches the speed of light at a critical radius:

R_LC = c / Ω = c P / (2π)

This is the light cylinder, a cylindrical surface coaxial with the rotation axis. Nothing material can co-rotate beyond it. That single geometric fact splits the magnetosphere in two. Field lines that leave the star and return to it while staying inside R_LC form the closed zone: the plasma trapped on them co-rotates rigidly, like a rotating dead weight. Field lines that would have to cross the light cylinder to close cannot — they are forced open, peeling away from the magnetic poles and streaming outward as the relativistic pulsar wind. The footpoints of these open lines on the stellar surface define the polar caps, small circular regions around the magnetic poles.

The polar cap is tiny. Its angular radius scales as θ_pc ≈ √(R/R_LC), so the cap radius is r_pc ≈ R √(R/R_LC) ≈ R √(2πR/cP). For a one-second pulsar with R = 12 km this is only about 190 m across — a patch the size of a few city blocks on the surface of the star. Everything energetic a pulsar does is funneled through that little cap and its open field lines.

The Goldreich-Julian density: how much plasma

How much charge does the star need to supply? Exactly enough to cancel the parallel electric field everywhere — to make E·B = 0 — because any leftover parallel field would just accelerate more charge until it was screened. Solving that condition gives the Goldreich-Julian charge density:

ρ_GJ = −2 ε₀ Ω·B        (SI)
     = −(Ω·B) / (2π c)   (Gaussian)

Converting to a number density of charges near the surface gives a famous estimate:

n_GJ ≈ 7 × 10⁻²  (B / 10¹² G) (1 s / P)  charges per cm³

Note the sign in ρ_GJ: because it depends on Ω·B, the required charge density flips sign across the magnetosphere. The star is wrapped in regions of opposite net charge — negative above one pole, positive over the equatorial belt, for a typical geometry. This charge separation is what carries the global current and what makes the gaps possible. And n_GJ is only the minimum: real pulsars manufacture far more plasma than this through pair cascades, with multiplicity factors of 10³–10⁶, and that excess is exactly what makes coherent radio emission possible.

Gaps, beams, and the lighthouse

If the star could always supply exactly ρ_GJ everywhere, the parallel electric field would be perfectly screened and nothing interesting would happen. But there are places where the charge supply falls short — near the polar cap where field lines are open and charges escape to infinity, along the boundary between open and closed lines, and in the outer magnetosphere near the light cylinder. In those gaps an unscreened parallel field survives and accelerates particles to enormous energies.

A particle accelerated through even a fraction of that 10¹⁶ V potential reaches TeV energies. Following the curved field lines, it radiates curvature gamma rays; those gammas pair-produce in the strong field; the pairs radiate more gammas; and a runaway electron-positron cascade fills the open-field bundle with dense plasma. This plasma, bunched and streaming along the curved open lines, radiates the famous coherent radio beam — a tight cone only a few degrees wide, aligned with the magnetic dipole axis.

Here is the part that produces the pulses. The magnetic axis is tilted relative to the spin axis by some inclination angle α. As the star rotates, the beam cone sweeps around the sky like the rotating lamp of a lighthouse. The pulsar shines continuously; you only detect it during the brief instant each rotation when the beam crosses your line of sight. Miss the beam and the star is radio-silent to you. Catch both poles and you see a main pulse plus an interpulse. The pulse you record is not the star blinking — it is geometry, the beam sweeping past Earth once per spin period.

Spin-down: the cost of shining

All of this energy comes from one place: the star's rotation. A rotating, inclined magnetic dipole radiates electromagnetic power and drives a magnetized wind, and both extract rotational kinetic energy. The classic vacuum-dipole spin-down luminosity is:

L = (2 / 3c³) B² R⁶ Ω⁴ sin²α

where B is the surface dipole field, R the stellar radius, Ω the spin rate, and α the inclination angle. Set this equal to the rate of loss of rotational energy, L = −I Ω Ω̇ (with moment of inertia I ≈ 10⁴⁵ g cm²), and you get the spin-down: the period P slowly lengthens, Ṗ > 0. Two beautiful diagnostics fall out of measuring P and Ṗ:

Surface field:        B ≈ 3.2 × 10¹⁹ √(P Ṗ)  gauss
Characteristic age:   τ = P / (2 Ṗ)

You read a pulsar's magnetic field and age straight off its timing, without ever resolving it as more than a point. The vacuum formula gets the scalings right but is known to be incomplete — it wrongly predicts that an aligned rotator (α = 0) does not spin down at all. Global force-free and particle-in-cell simulations (Contopoulos, Kazanas & Fendt 1999; Spitkovsky 2006) corrected this: even an aligned pulsar spins down, because the open-field current and the wind carry energy regardless of α. The modern form is roughly L ≈ (μ²Ω⁴/c³)(1 + sin²α), about 50% larger than vacuum for the aligned case.

Worked example: the Crab pulsar

The Crab pulsar (PSR B0531+21), the spinning remnant of the supernova seen in 1054 AD, is the textbook case. Its parameters:

Period         P  = 33.5 ms = 0.0335 s
Period deriv.  Ṗ  = 4.2 × 10⁻¹³ s/s
Moment of I    I  ≈ 10⁴⁵ g cm²

Light cylinder. Ω = 2π/P = 188 rad/s, so

R_LC = c / Ω = (3.0 × 10⁸ m/s) / (188 s⁻¹) ≈ 1.6 × 10⁶ m ≈ 1600 km

The Crab's entire co-rotating magnetosphere fits inside a cylinder 1600 km in radius — only about 130 stellar radii. Compare a slow 1-second pulsar: R_LC ≈ 48,000 km. Or a 1.4-ms millisecond pulsar: R_LC ≈ 67 km, barely five stellar radii, which is why millisecond pulsars have the most compact, intense magnetospheres known.

Surface field. B ≈ 3.2 × 10¹⁹ √(P Ṗ) = 3.2 × 10¹⁹ √(0.0335 × 4.2 × 10⁻¹³) ≈ 3.8 × 10¹² gauss — a few teratesla in SI, entirely normal for a young pulsar.

Characteristic age. τ = P/(2Ṗ) = 0.0335 / (2 × 4.2 × 10⁻¹³) ≈ 4.0 × 10¹⁰ s ≈ 1260 years. The true age is 972 years (1054 → 2026), so τ overestimates it by ~30%, exactly the expected level of agreement given the simplifying assumptions.

Spin-down luminosity. L = −I Ω Ω̇. With Ω̇ = −2π Ṗ/P² this works out to L ≈ 4.5 × 10³¹ W ≈ 4.5 × 10³⁸ erg/s — roughly the output of 100,000 Suns, almost all of it going into the surrounding Crab Nebula rather than into the radio beam. The period lengthens by about 36 nanoseconds per day. Almost nothing escapes as radio: the radio luminosity is a part in 10⁵ of the spin-down power.

Regimes and zoo of magnetospheres

Class	Period P	B (gauss)	R_LC	Powered by	Example
Young rotation-powered	10–100 ms	10¹²–10¹³	~10³ km	spin-down	Crab, Vela
Normal pulsar	0.1–10 s	10¹¹–10¹³	10⁴–10⁵ km	spin-down	PSR B0329+54
Millisecond pulsar	1.4–10 ms	10⁸–10⁹	67–480 km	spin-down (recycled)	PSR B1937+21
Magnetar	2–12 s	10¹⁴–10¹⁵	~10⁵ km	magnetic decay	SGR 1806−20
Death-line pulsar	5–8 s	~10¹¹	~3×10⁵ km	marginal pair supply	PSR J2144−3933
Pulsar wind nebula host	10–500 ms	10¹²–10¹³	~10³–10⁴ km	wind termination shock	Crab Nebula engine

The bottom-left of the period–period-derivative (P–Ṗ) diagram is bounded by the death line: when the period is long enough and the field weak enough, the polar-cap voltage drops below the threshold for pair production, the cascade shuts off, the plasma multiplicity collapses to ~1, and the radio beam switches off. PSR J2144−3933, with P = 8.5 s, sits suspiciously far past most theoretical death lines and remains a puzzle. Magnetars, at the opposite extreme, are powered not by spin-down but by the decay of their colossal 10¹⁴–10¹⁵ G fields, and their magnetospheres are twisted and current-loaded rather than the clean dipole-plus-wind picture above.

Quantitative analysis: the force-free magnetosphere

Why can we get away with simple charge-density arguments at all? Because the magnetic energy density near a pulsar dwarfs everything else — the plasma's rest-mass and thermal energy are negligible by many orders of magnitude. The plasma therefore has no dynamical say: it cannot push back on the field. This is the force-free limit, in which the total Lorentz force density on the plasma vanishes:

ρ E + (J × B) / c = 0

The Goldreich-Julian density is the leading-order solution of this equation for a rotating dipole. But the full global structure — where the field lines actually go, how much current flows, exactly how fast the star spins down — requires solving the force-free equations numerically. Contopoulos, Kazanas & Fendt (1999) found the first self-consistent steady-state aligned solution, revealing three key features: a smooth open-field bundle anchored on the polar cap, a Y-shaped current sheet that begins at the light cylinder and extends outward in the equatorial plane, and a spin-down rate about 50% higher than the vacuum dipole even with no inclination.

Time-dependent simulations (Spitkovsky 2006; Kalapotharakos & Contopoulos 2009) extended this to oblique rotators, and modern global particle-in-cell models (Cerutti, Philippov, Spitkovsky) self-consistently create the pairs, populate the gaps, and let the high-energy emission emerge rather than be assumed. A robust result across these models is that much of the gamma-ray emission — the kind Fermi-LAT has detected from over 290 pulsars — originates not at the polar cap but near the current sheet just beyond the light cylinder, a substantial revision of the original polar-cap picture. Where the radio coherent emission is generated within this structure remains the hardest open problem in the field.

Observational status and applications

• Radio timing. Measuring P and Ṗ for the ~3400 known pulsars (ATNF catalogue) yields each star's inferred field and age via B ≈ 3.2×10¹⁹√(PṖ) and τ = P/(2Ṗ). The P–Ṗ diagram is the H-R diagram of pulsar astrophysics.
• Gamma-ray gaps. The Fermi Large Area Telescope has detected pulsed GeV emission from 290+ pulsars; the double-peaked gamma-ray light curves and the radio-to-gamma phase lag are the cleanest probes of where the accelerating gaps sit relative to the light cylinder.
• Pulsar wind nebulae. The spin-down luminosity reappears as synchrotron and inverse-Compton emission when the wind shocks against the supernova remnant — the Crab Nebula radiates ~5×10³¹ W siphoned from its pulsar's rotation.
• Millisecond-pulsar clocks. Because their compact magnetospheres are extraordinarily stable, recycled millisecond pulsars keep time to nanosecond precision over years, enabling pulsar timing arrays to search for nanohertz gravitational waves.
• Magnetar bursts. Magnetar magnetospheres, twisted and current-loaded, episodically reconfigure, releasing the giant flares and short gamma-ray bursts that make them among the most violent objects in the Galaxy.

Common pitfalls and misconceptions

• "The pulsar turns on and off." No — it emits continuously. You see a pulse only when the lighthouse beam sweeps across your line of sight. The pulsing is geometry, not a flickering source.
• "A pulsar's magnetosphere is a vacuum dipole." The whole point of Goldreich-Julian is that it cannot be empty. A vacuum would have an unscreened 10¹⁶ V parallel field that immediately fills the space with plasma.
• "Aligned rotators don't spin down." The vacuum-dipole formula's sin²α factor says so, but it is wrong. Force-free and PIC simulations show an aligned pulsar still loses energy through its open-field current and wind.
• "The light cylinder is a physical wall." It is a geometric surface, not a barrier. Plasma and field lines pass through it freely; what cannot happen is rigid co-rotation beyond it.
• "The radio beam carries most of the energy." It carries a part in ~10⁵ of the spin-down power. The bulk goes into the high-energy emission and especially the relativistic wind feeding the nebula.
• "Surface field B from timing is the true field everywhere." B ≈ 3.2×10¹⁹√(PṖ) is the dipole field at the equator inferred under vacuum-dipole assumptions; the true field can have higher-order multipoles, especially near the surface, and magnetar fields are even more distorted.