Astronomy & Celestial Mechanics
the sky, computed
Orbits

A resonance with the planet itself

Every other resonance in celestial mechanics is a commensurability between two orbits. A charged dust grain has a third clock available — the planet's rotation, which sweeps its magnetic field past the grain — and the commensurability with that selects by charge-to-mass ratio, which means by grain size.

Assumes Planetary rings and Resonance.

18 min read 10 figures Small perturbations, large consequencesTwo bodies, and no more

Jupiter’s main ring is a narrow band of dust between 1.72 and 1.81 planetary radii. Inside it is something stranger: the halo, a cloud of fine material extending from about 1.71 radii inward to about 1.40, and puffed some ten thousand kilometres out of the ring plane — a torus rather than a disc, in a system where everything else is flat.

The halo’s two edges are the subject of this essay, because they are not where any moon puts them. They are where a resonance with the planet’s rotation puts them.

A resonance with the planet's rotation, and it sorts by size. Above: the radii at which a charged grain's orbital frequency is commensurate with the planet's spin, in planetary radii, with the measured edges of Jupiter's halo ring drawn over them. The synchronous radius is 2.24; the 3:2 resonance falls at 1.71 and the 2:1 at 1.41. The halo's outer boundary is at 1.71 and its inner extent near 1.4 — the ring ends where the resonances are, and both numbers were measured by a spacecraft camera with no reference to this arithmetic. Below: which grains care. The charge on a grain is proportional to its radius and the mass to the cube, so the force per unit mass goes exactly as the inverse square of the size, and the resonance grips sub-micron dust while leaving anything larger on a Keplerian orbit. That is why the halo is a cloud of fine dust puffed a thousand kilometres out of the ring plane while the coarse material stays flat: the same field acting on the same orbit sorts the material by size, which no gravitational resonance can do.
Fig. 1 Above: the radii at which a charged grain’s orbital frequency is commensurate with Jupiter’s spin, with the measured halo edges drawn over them. Below: which grains care — the force per unit mass goes exactly as the inverse square of the grain radius, so the resonance grips sub-micron dust and leaves anything larger on a Keplerian orbit.

A grain is not a rock

Everything in celestial mechanics up to this point has assumed that the only force on a body is gravity. For a grain a micron across in a planetary magnetosphere that assumption fails.

Dust in a plasma charges. Electrons, being more mobile, strike a grain more often than ions do, so an isolated grain in a plasma charges negatively until the resulting potential repels enough electrons to balance the currents — typically a few volts, tens of volts in an energetic plasma, and positive rather than negative where photoemission from sunlight dominates.

A charged grain moving through a magnetic field feels a Lorentz force. The relevant quantity is the force per unit mass, which is the charge-to-mass ratio times the field times the velocity.

The charge on a grain is proportional to its radius, because it is a capacitor. The mass is proportional to the cube. So the charge-to-mass ratio goes as the inverse square of the radius, exactly, and the ratio of Lorentz force to gravity does the same.

That exponent is the whole reason the effect sorts material. A grain of a tenth of a micron feels a hundred times the specific Lorentz force of a one-micron grain, and ten thousand times that of a ten-micron one. There is no gradual transition: dust below a threshold size is dominated by electromagnetic forces and dust above it is not.

It is worth setting the threshold against something. For a grain in Jupiter’s inner magnetosphere at a few volts of potential the Lorentz and gravitational forces are comparable at a radius of roughly a tenth of a micron. Below that the grain is essentially a heavy ion and is swept out of the system in months; above about a micron it is essentially a rock. The window in which a grain is a dynamical hybrid — genuinely orbiting, and genuinely perturbed — is less than a decade wide in size, and it is exactly the material the halo is made of.

The charging itself is not a constant, which complicates matters and produces some of the interesting behaviour. A grain passing into a planet’s shadow loses its photoelectron current and its potential changes within minutes; a grain crossing a plasma boundary charges to a different equilibrium. So the charge-to-mass ratio is a function of position around the orbit, and a periodically varying charge is another route to a resonance entirely — a gyrophase drift, whose treatment requires the same machinery as a slowly varying orbital element.

The third clock

An orbiting grain sees a magnetic field that is not static, because the planet’s field is not aligned with its rotation axis and the planet rotates.

Jupiter’s dipole is tilted about ten degrees from its spin axis and offset from its centre. To a grain in a circular orbit, the field therefore oscillates, at a frequency equal to the difference between the planet’s rotation rate and the grain’s own orbital rate.

That is a periodic forcing, and a periodic forcing of an orbit produces a resonance whenever its frequency matches a natural frequency of the orbit — which for a nearly circular, nearly equatorial orbit is the orbital frequency itself — the same condition any resonance in this collection is built on.

The condition works out to a commensurability between the orbital frequency and the planet’s rotation frequency in the ratio of small integers. The radii at which it is satisfied follow immediately from Kepler’s third law: they are the synchronous radius multiplied by the ratio to the power two thirds.

For Jupiter the synchronous radius is 2.24 planetary radii. The three-to-two resonance, where a grain orbits three times per two rotations, sits at 1.71; the two-to-one at 1.41.

The synchronous radius deserves a note of its own, because it divides the system. Outside it a grain orbits more slowly than the field sweeps past, and the drag from the corotating plasma pushes it outward; inside it the grain overtakes nothing and is pushed inward. Jupiter’s main ring and halo are both inside synchronous, which is why the material there drifts toward the planet and is lost rather than accumulating.

Saturn is the instructive contrast. Its rotation is slower and its rings are far more extensive, so its synchronous radius at 1.87 planetary radii falls inside the C ring — with most of the ring system outside it. The dust dynamics of the two systems therefore differ in sign over most of their extent, and the structures produced are correspondingly different.

A resonance with the planet's rotation, and it sorts by size. Above: the radii at which a charged grain's orbital frequency is commensurate with the planet's spin, in planetary radii, with the measured edges of Jupiter's halo ring drawn over them. The synchronous radius is 1.86; the 3:2 resonance falls at 1.42 and the 2:1 at 1.17. The halo's outer boundary is at 1.42 and its inner extent near 1.18 — the ring ends where the resonances are, and both numbers were measured by a spacecraft camera with no reference to this arithmetic. Below: which grains care. The charge on a grain is proportional to its radius and the mass to the cube, so the force per unit mass goes exactly as the inverse square of the size, and the resonance grips sub-micron dust while leaving anything larger on a Keplerian orbit. That is why the halo is a cloud of fine dust puffed a thousand kilometres out of the ring plane while the coarse material stays flat: the same field acting on the same orbit sorts the material by size, which no gravitational resonance can do.
Fig. 2 The same arithmetic run for Saturn, which is the check that it is arithmetic. Nothing changes but the mass, the radius and the rotation period; the resonances fall out at 1.42 and 1.18 planetary radii, against Jupiter’s 1.71 and 1.41, purely because Saturn turns more slowly and is less dense. Both lie inside the C ring’s inner edge, in a region that holds almost no material — so Saturn has the resonances and not the halo, and the reason is that its dust source is somewhere else entirely. A mechanism that predicts a structure where there is material and predicts nothing where there is none is being tested by the second case as much as by the first.
A straight line whose slope is a mass per unit area. The measurement that the previous picture is the raw material for. Each point is one crest of the density wave: its number outwards from the resonance along the horizontal axis, and the square of its distance from the resonance up the vertical one. The points lie on a straight line through the origin to within 7.1e-14 per cent, which is what the phase accumulating as the square of the distance means, and the slope of that line is 8π²Gσr_L/3(m−1)Ω_L² — one expression containing one unknown. Solving it gives 35.0 kilograms per square metre. Everything else in the expression is known to several figures from the moon's orbit and the planet's mass, so the error on the ring's mass is essentially the error on reading crest positions off a light curve. A ring a few tens of metres thick, spread over an area larger than the Earth, is weighed by fitting a line to twenty-six numbers.
Fig. 3 The dispersion relation the resonance launches a wave into. A differentially rotating self-gravitating sheet supports density waves whose propagation speed depends on the surface density and the epicyclic frequency — so a resonance does not simply clear a gap here, it radiates. Which of the two happens depends on whether the ring can carry a wave away from the resonance faster than the torque piles material up at it, and the answer is a property of the ring rather than of the moon.

What the resonance does

A resonance pumps something, and here it pumps inclination as well as eccentricity.

The Lorentz force on a grain in an inclined orbit has a component perpendicular to the ring plane, oscillating at the beat frequency. At a resonance the driving is in phase with the grain’s vertical motion cycle after cycle, and the inclination grows.

The growth is not unbounded — as the inclination increases the grain samples different field strengths and drifts out of resonance — but it can reach tens of degrees, which for material at 1.5 planetary radii is a vertical excursion of thousands of kilometres.

A straight line whose slope is a mass per unit area. The measurement that the previous picture is the raw material for. Each point is one crest of the density wave: its number outwards from the resonance along the horizontal axis, and the square of its distance from the resonance up the vertical one. The points lie on a straight line through the origin to within 7.1e-14 per cent, which is what the phase accumulating as the square of the distance means, and the slope of that line is 8π²Gσr_L/3(m−1)Ω_L² — one expression containing one unknown. Solving it gives 70.0 kilograms per square metre. Everything else in the expression is known to several figures from the moon's orbit and the planet's mass, so the error on the ring's mass is essentially the error on reading crest positions off a light curve. A ring a few tens of metres thick, spread over an area larger than the Earth, is weighed by fitting a line to twenty-six numbers.
Fig. 4 The alternative outcome, at twice the surface density. A resonance in a massive ring launches a wave rather than pumping individual particles, because the ring’s own self-gravity supplies a restoring force and the disturbance propagates away as fast as it is driven. Doubling the surface density steepens the dispersion relation and the wave carries the angular momentum off more effectively still. That is the whole reason the halo is possible: a dusty ring has a surface density orders of magnitude below this, so there is no collective response at all, and the torque acts on each grain separately according to its own charge-to-mass ratio. Sorting requires a ring too thin to respond as a fluid.

That is the halo. Material drifting inward from the main ring reaches the three-to-two resonance at 1.71 radii, has its inclination pumped, and leaves the ring plane. The halo’s outer edge is therefore not a boundary of the material’s distribution in radius but the radius at which the material stops being flat.

The inner edge at 1.40 is the two-to-one resonance and its role is less clean. Below it the grains’ orbits become sufficiently perturbed that they are lost — into the planet’s atmosphere, or by further inward drift — and the halo’s brightness falls away. Loss into the planet is not hypothetical: Jupiter’s upper atmosphere shows a persistent ring of infrared emission at the latitude to which the halo’s field lines connect.

The vertical structure is the part that a photograph makes obvious and that no gravitational mechanism could produce. Ring material is flat because collisions damp vertical motion: a particle on an inclined orbit crosses the ring plane twice per revolution, collides, and loses its vertical velocity. Flatness is a consequence of collisions, so a vertically extended dust population is one in which collisions are too rare to damp what is being pumped — which for a halo of sub-micron grains at that density they are. That is the same competition between damping and driving that decides whether an eccentricity survives, transposed from a binary orbit to a ring.

Why the material is drifting inward at all

The picture requires a supply and a drift, and both are worth stating.

The supply is impacts. Micrometeoroids strike the small inner moons Metis and Adrastea and eject material at speeds above those bodies’ tiny escape velocities, so the moons act as continuous dust sources. That is the standard account of every dusty ring in the solar system, and the ring’s radial extent matches the moons’ orbits closely — the same relationship a shepherded edge has to the moon that maintains it, with a source in place of a boundary. The drift is Poynting–Robertson drag plus plasma drag. A grain absorbing sunlight and re-radiating it isotropically in its own frame loses angular momentum, and a grain moving relative to a corotating plasma is dragged by it. Inside the synchronous radius the plasma corotates faster than the grain orbits, so plasma drag pushes it outward; outside, inward. That is one of the reasons the inner ring system’s dynamics are different from the outer.

Timescales are short. A one-micron grain at two Jovian radii drifts appreciably in years to decades, which is nothing, and the whole halo is therefore a steady state maintained by continuous resupply rather than a structure that has been there since the system formed.

An edge where two torques balance, 21 kilometres from the shepherd. Two torques on the edge of a ring, against distance from a shepherding moon, both axes logarithmic and both scaled by the same combination of surface density, radius and orbital rate so that only their shapes are being compared. The flat line is the viscous torque, which comes from collisions between ring particles and does not care how far away anything is; it is drawn for a kinematic viscosity of 12 square centimetres a second, within the range ring seismology gives. The falling line is the moon's, summed over the first-order resonances that crowd together as the gap narrows, which makes it an inverse cube. A flat curve and an inverse cube cross once, and the crossing is where an edge can sit: closer in the moon wins and pushes the material back, further out viscosity wins and the ring spreads. For Daphnis and the Keeler gap the balance lands 20.8 kilometres out against a measured half-width of 21, which is agreement to well inside the uncertainty on the viscosity — and it is the only handle anybody has on that viscosity, since the quantity being inferred is the collision rate among particles a metre across, a billion kilometres away.
Fig. 5 The gravitational way of holding a ring edge in place, for contrast with the electromagnetic one. A shepherd moon’s torque balances the ring’s own viscous spreading, and the edge sits where the two are equal — twenty-one kilometres from the moon in the case drawn. Every particle at that radius is treated identically, whatever its size, because gravity is a force per unit mass. The halo’s edges are the opposite kind of boundary: they are where a subset of the material stops behaving like the rest, and the material a micron across sails through the same radius unaffected. A balance of torques makes an edge; a resonance that sorts makes two populations.

That is the general condition of every dusty ring, and it is the reason dusty rings are informative. A structure with a lifetime of decades is a structure whose present appearance is set entirely by present conditions, so it reports the current field, the current plasma and the current impact rate — where a massive ring’s structure records processes lasting millions of years.

What makes the argument checkable

Two numbers were predicted and then measured, and that is what distinguishes this from a plausible story.

The resonance radii follow from the planet’s mass and its rotation period, both of which are known to many significant figures, through Kepler’s third law and a ratio of integers. There is nothing to adjust: three to two gives 1.71 radii, two to one gives 1.41. The halo’s edges were measured by spacecraft imaging: the outer boundary at about 1.71 and the material fading out below about 1.40. The agreement is to the accuracy of the imaging.

Ring edges where a moon's arithmetic says they should be. Saturn's main rings drawn to scale in radius, with every resonance the five nearest moons place inside them. A resonance radius is arithmetic on one number: a ring particle completing p orbits while the moon completes q sits at a(q/p)^(2/3), and the moon's semi-major axis is known from tracking to a few kilometres. The three marked features are the test. The outer edge of the B ring, mapped at 117,580 km by watching stars set behind it, is predicted at 116,882 by Mimas's 2:1 — an error of 0.59 per cent. The outer edge of the A ring at 136,775 is predicted at 136,668 by Janus's 7:6. The Mimas 5:3 density wave, which is a wave rather than an edge, is predicted at 131,988. Every one of the three is placed by a body tens of thousands of kilometres away that has never been anywhere near the ring, and the worst of the three is out by 0.59 per cent. What the picture cannot show is why some resonances make a sharp edge, some a wave and most nothing at all — that depends on the resonance's order and on whether the ring has enough mass there to carry a wave.
Fig. 6 The Mimas 2:1 on its own, which is the comparison the previous paragraph implies. The Cassini division’s inner edge sits at that resonance and was computed the same way — a ratio of small integers and Kepler’s third law — so the two arguments have identical form and differ only in what the second body is. The distinction that matters is what happens to a particle that is not the size the mechanism selects. At the Mimas resonance there is no such particle: a boulder and a grain are cleared alike. At a Lorentz resonance the grain leaves and the boulder does not, and the structure is a sorted population rather than a gap.
Ring edges where a moon's arithmetic says they should be. Saturn's main rings drawn to scale in radius, with every resonance the five nearest moons place inside them. A resonance radius is arithmetic on one number: a ring particle completing p orbits while the moon completes q sits at a(q/p)^(2/3), and the moon's semi-major axis is known from tracking to a few kilometres. The three marked features are the test. The outer edge of the B ring, mapped at 117,580 km by watching stars set behind it, is predicted at 116,882 by Mimas's 2:1 — an error of 0.59 per cent. The outer edge of the A ring at 136,775 is predicted at 136,668 by Janus's 7:6. The Mimas 5:3 density wave, which is a wave rather than an edge, is predicted at 131,988. Every one of the three is placed by a body tens of thousands of kilometres away that has never been anywhere near the ring, and the worst of the three is out by 0.59 per cent. What the picture cannot show is why some resonances make a sharp edge, some a wave and most nothing at all — that depends on the resonance's order and on whether the ring has enough mass there to carry a wave.
Fig. 7 The ring system’s other resonant structures for comparison. Gaps and edges maintained by moons occur at radii computed the same way and checked the same way, and the Lorentz resonances sit among them as a family with a different second body.

The grain size is the second prediction. If the effect is electromagnetic and the charge-to-mass ratio goes as the inverse square of the size, then only the finest material should be affected — and photometry of the halo, from its colour and its scattering behaviour, gives grain sizes below about half a micron, against the several microns of the main ring.

A ring sorted by size, with the fine material in a torus and the coarse in a sheet, at radii predicted by an integer ratio. That is a strong result for a subject where most structures are attributed to moons that have not been found.

What was known before the spacecraft

The halo was not predicted. It was photographed by Voyager in 1979, at a time when Jupiter was not known to have rings at all until the same mission found them, and the explanation followed.

That order of events is worth stating because it is the usual one in this part of the subject and it sets the standard for what counts as a confirmation. A theory constructed after the observation can always be made to fit; what makes the Lorentz resonance account convincing is that it fits with no adjustable parameters, and that it made a further prediction — about grain size — that was tested afterwards. The size prediction is the one that mattered. Photometry of the halo across several wavelengths and scattering angles gives a size distribution, because small grains scatter forward preferentially and large ones do not, and the halo’s strong forward-scattering says its material is well below a micron. The main ring, measured the same way, is coarser. Two populations at adjacent radii with different size distributions, separated at exactly the radius where the electromagnetic force becomes competitive for one and not the other.

There is a residual difficulty that honest accounts include. The pumping at a resonance takes many orbits, and a grain drifting inward passes through the resonance in a time that depends on the drift rate — so whether it is pumped at all is a question of how slowly it crosses. Calculations give the right order and the halo’s precise vertical profile is not reproduced in detail, which is where the modelling effort currently sits.

Where else it operates

The same physics applies wherever there is dust, a magnetic field and rotation, and it does not always produce a halo.

Saturn’s spokes are the most famous case: transient radial markings on the B ring, seen by two flybys and then by an orbiting spacecraft, appearing and dissipating in hours. They are generally attributed to fine grains lifted electrostatically above the ring plane, and their appearance correlates with the ring’s orientation relative to the Sun in a way that supports a charging mechanism.

Jupiter’s gossamer rings, outside the main ring, are shaped by the inclinations of their source moons rather than by Lorentz resonances, which is a useful negative: dusty structures are not all magnetic, and the diagnostic is whether the vertical extent matches a source’s inclination or a resonance’s radius.

A wave whose crests count out 35 kilograms per square metre of ring. Optical depth across a spiral density wave in Saturn's A ring, drawn outwards from the 7:6 inner Lindblad resonance with Janus at 136,668 km. The wave is launched at the resonance on the left and damps away to the right. Its wavelength is not constant: it starts at about 6.0 km and has shortened to 1.30 km by the last crest drawn, because the wavenumber grows in proportion to the distance from resonance and the ring's own self-gravity is the only restoring force in the dispersion relation. That makes the pattern a chirp with exactly one unknown in it. Fitting the 21 crest positions actually drawn here — the square of each one's distance from resonance against its number, which is a straight line — returns a surface density of 35.0 kilograms per square metre against the 35 the profile was built from. The rings have been weighed this way rather than by anything touching them: the mass per unit area follows from counting bright bands in a light curve as a star sets behind the ring.
Fig. 8 The Janus 7:6, at the A ring’s outer edge, as the third member of that family. It is a density wave with counted crests, and the count is a surface density — thirty-five kilograms per square metre, weighed by watching a wave cross the material. Set the three side by side and the diagnostic in the paragraph above becomes a general one: a gravitational resonance in a massive ring gives a wave whose wavelength measures the mass; the same resonance in a tenuous ring gives an edge; and a resonance with the planet’s own rotation gives neither, because it acts on a property no wave can average over. What a resonance does to a ring is a statement about the ring, not about the resonance.

And in a protoplanetary disc the same charge-to-mass argument decides whether the smallest grains are coupled to the magnetic field, which affects how the gas is coupled to it too — grains soak up free electrons, so their abundance and size distribution control the ionisation and therefore whether the disc’s turbulence operates at all.

Saturn’s spokes, in more detail

The other place where dust, charge and rotation combine to produce something a purely gravitational account cannot is worth setting out, because it is the case where the mechanism is least settled.

Spokes are radial markings on Saturn’s B ring, dark in backscattered light and bright in forward-scattered light — which says immediately that they are made of grains far smaller than the ring particles, since the scattering behaviour reverses only for particles comparable to the wavelength.

They form in minutes, extend over thousands of kilometres radially, and shear out over hours as differential rotation acts on them. Their radial extent and their sharp appearance near the synchronous radius are the clues: a radial feature in a differentially rotating disc is a feature that was created faster than the disc could shear it, which bounds the formation time.

A wave whose crests count out 35 kilograms per square metre of ring. Optical depth across a spiral density wave in Saturn's A ring, drawn outwards from the 5:3 inner Lindblad resonance with Mimas at 131,988 km. The wave is launched at the resonance on the left and damps away to the right. Its wavelength is not constant: it starts at about 5.6 km and has shortened to 1.13 km by the last crest drawn, because the wavenumber grows in proportion to the distance from resonance and the ring's own self-gravity is the only restoring force in the dispersion relation. That makes the pattern a chirp with exactly one unknown in it. Fitting the 24 crest positions actually drawn here — the square of each one's distance from resonance against its number, which is a straight line — returns a surface density of 35.0 kilograms per square metre against the 35 the profile was built from. The rings have been weighed this way rather than by anything touching them: the mass per unit area follows from counting bright bands in a light curve as a star sets behind the ring.
Fig. 9 The ring’s ordinary structure for contrast. A density wave launched at a moon’s resonance is a coherent, long-lived pattern whose wavelength encodes the surface density; a spoke is a transient that lasts a few hours and encodes nothing about the ring’s mass at all.

The favoured account is electrostatic levitation: a sudden ionisation event above the ring — a meteoroid impact producing a plasma cloud, or an electron beam from the magnetosphere — charges the surface grains sufficiently that they lift off electrostatically. Whether the trigger is impacts is not established, but the seasonal behaviour is: spokes vanish when the rings are near their maximum tilt to the Sun and return near equinox, which is what a photoelectric charging balance would predict.

That seasonal dependence is a good example of a prediction whose test required waiting. It was noticed in the flyby data, explained, and then checked fifteen years later when an orbiting spacecraft was present through a full change of season.

What this rung adds to the ladder

The other resonances in this collection are between two orbits: a moon and a ring particle, two moons, a planet and an asteroid. The forcing is gravitational, the strength depends on a mass ratio, and every body of every size at the resonant radius responds identically — which is why a resonance can clear a gap rather than sorting one.

Here the second body is the planet’s own rotation, the forcing is electromagnetic, and the strength depends on a property of the individual grain. A resonance that selects by size is a different kind of object, and it produces structures that a gravitational resonance cannot: a population sorted rather than cleared. Sorting is what makes it a diagnostic: the presence of a size-sorted structure is evidence that a field is acting, in a system where the field itself is measured only by a spacecraft flying through it.

Ring edges where a moon's arithmetic says they should be. Saturn's main rings drawn to scale in radius, with every resonance the five nearest moons place inside them. A resonance radius is arithmetic on one number: a ring particle completing p orbits while the moon completes q sits at a(q/p)^(2/3), and the moon's semi-major axis is known from tracking to a few kilometres. The three marked features are the test. The outer edge of the B ring, mapped at 117,580 km by watching stars set behind it, is predicted at 116,882 by Mimas's 2:1 — an error of 0.59 per cent. The outer edge of the A ring at 136,775 is predicted at 136,668 by Janus's 7:6. The Mimas 5:3 density wave, which is a wave rather than an edge, is predicted at 131,988. Every one of the three is placed by a body tens of thousands of kilometres away that has never been anywhere near the ring, and the worst of the three is out by 0.59 per cent. What the picture cannot show is why some resonances make a sharp edge, some a wave and most nothing at all — that depends on the resonance's order and on whether the ring has enough mass there to carry a wave.
Fig. 10 Where the strongest resonances of the inner moons fall, and what sits at them. The A ring’s outer edge is at the Janus 7:6 and the Cassini division begins at the Mimas 2:1 — neither a coincidence, both a torque acting at a commensurability. But a resonance with the planet itself is a different animal: the Lorentz resonances that shape the faint dusty rings come from the planet’s own rotating magnetic field acting on charged grains, and they fall where the field’s rotation period is commensurate with the orbit rather than where a moon is.

It also breaks an assumption that has held everywhere else in the collection. Up to this point a body’s trajectory has depended only on where it is and how fast it is going, never on what it is made of or how large it is. A charged grain’s orbit depends on both, and once that is true the phrase “the orbit at this radius” stops meaning anything: there are as many orbits as there are grain sizes.

That is worth carrying forward, because it is the point at which celestial mechanics stops being a subject about point masses. The bodies are still small compared with their orbits, but they are no longer characterised by mass alone — and everything that follows about dust, about rings and about the material that fails to become planets happens in that regime.

The break with point-mass dynamics is worth stating once more in the form it takes here, because it is sharper than the usual one. A non-gravitational force that depends on a body’s properties is not new in this collection: radiation pressure, drag and thermal recoil all do. What is new is that this one produces a resonance, and a resonance is a phenomenon of the unperturbed problem’s frequencies rather than of the perturbation’s magnitude.

So the size dependence does not merely shift a trajectory by an amount proportional to the force; it decides whether a body is captured into a state qualitatively different from the one beside it. A grain of 0.4 microns and one of 0.6 at the same radius are on the same orbit until the resonance is reached, at which point one leaves the ring plane by thousands of kilometres and the other does not.

That is a sorting mechanism with a sharp threshold, built from a force nobody can measure directly, acting on a population nobody can resolve. The evidence that it operates is entirely circumstantial and entirely convincing: the ring ends where the arithmetic says, the material there is the size the arithmetic requires, and the vertical structure is the one the mechanism produces and no other mechanism does. Three independent agreements from one integer ratio and a charge-to-mass scaling is as much as a dust dynamicist is ever likely to get. The material cannot be sampled, the field is measured only where a spacecraft has flown, and the grains themselves are below the resolution of every camera that has looked. What is left is an integer, a scaling and two measured edges. That is a thin evidential base for a mechanism, and it is thicker than most of what planetary ring dynamics rests on. The two effects together mark the boundary of the classical subject. Above about a kilometre a body is a point mass and its orbit is a solved problem; below a micron it is a charged particle and belongs to plasma physics; and in the eight decades between, its trajectory depends on its size, its spin, its albedo, its conductivity and its surface potential. Almost all the mass in a debris disc is at the top of that range and almost all the area, which is what is seen, is near the bottom.

Where this goes next

The Kirkwood gaps. Asteroid numbers against semi-major axis, with the resonant radii marked. Each gap sits where the orbital period is a simple fraction of Jupiter's, and each of those radii is computed from the harmonic law rather than placed by eye. Resonance clears a gap in one place and locks a moon in another When two orbital periods are in a simple ratio, small tugs stop averaging away and start accumulating. Sometimes that empties a region entirely. Sometimes it holds three moons together for the age of the solar system. An edge where two torques balance, 21 kilometres from the shepherd. Two torques on the edge of a ring, against distance from a shepherding moon, both axes logarithmic and both scaled by the same combination of surface density, radius and orbital rate so that only their shapes are being compared. The flat line is the viscous torque, which comes from collisions between ring particles and does not care how far away anything is; it is drawn for a kinematic viscosity of 12 square centimetres a second, within the range ring seismology gives. The falling line is the moon's, summed over the first-order resonances that crowd together as the gap narrows, which makes it an inverse cube. A flat curve and an inverse cube cross once, and the crossing is where an edge can sit: closer in the moon wins and pushes the material back, further out viscosity wins and the ring spreads. For Daphnis and the Keeler gap the balance lands 20.8 kilometres out against a measured half-width of 21, which is agreement to well inside the uncertainty on the viscosity — and it is the only handle anybody has on that viscosity, since the quantity being inferred is the collision rate among particles a metre across, a billion kilometres away. An edge is a balance, not a boundary A ring of particles spreads, because collisions move angular momentum outward. Something has to push back, and the something is a small moon whose torque falls as the inverse cube of the gap. A flat curve and an inverse cube cross once, and the crossing is the sharp edge — which is why an edge exists at all rather than a gradient. Three motions, three decades apart, and that is the point. The three periods of a trapped 1-MeV electron's motion against the shell it is trapped on, on a logarithmic time axis. It gyrates about a field line in 0.22 milliseconds, bounces between its mirror points in 0.33 seconds, and drifts right round the planet in 16 minutes — ratios of 1498 and 3015 at L = 4. Each motion carries a conserved quantity: the magnetic moment, the longitudinal invariant, and the magnetic flux the drift shell encloses. The separation is what makes them conserved. An invariant survives anything that changes slowly compared with its own period, so a disturbance lasting minutes destroys the third and leaves the first two untouched — and a particle that keeps its magnetic moment while being moved inward to a stronger field must gain energy. That is not a loophole; it is how the belts are filled. Three clocks and nothing to fall onto A charged particle in a dipole field gyrates, bounces and drifts, on timescales a millisecond, a second and a quarter of an hour. The three periods are three decades apart, and that separation is not a curiosity — it is the reason each motion has a conserved quantity, and the reason a magnetic storm can accelerate particles rather than merely stir them.

The objects this essay names

Each one links to every other essay that touches it.

Charge to mass ratioDust dynamicsGrain chargingGyrophase driftJupiter haloLorentz resonanceMagnetospherePlanetary ringsPoynting robertson dragResonanceSynchronous orbit