Spiral Density Waves

A galactic disk rotates differentially: the circular angular speed $\Omega(R)$ falls with radius, so any material feature — a chain of stars painted along a spiral arm — is sheared out within a few orbital periods. This is the winding problem, and it is fatal for the naive picture. If arms were fixed collections of stars, a flat rotation curve would coil them from open to tight in about $10^9$ years, yet the sky is full of open, few-armed grand-design spirals that look dynamically old — something is wrong with treating an arm as a thing made of matter. The density-wave hypothesis, argued by Bertil Lindblad and made quantitative by C. C. Lin and Frank Shu in 1964, resolves the paradox with a single reframing: the spiral is not an object but a wave — a self-consistent overdensity in surface density and gravitational potential that rotates rigidly at one pattern speed $\Omega_p$ , independent of $R$ . Stars and gas orbit faster than the pattern inside corotation and slower outside; they stream through the arm the way cars pile into a slow-moving traffic jam and thin out beyond it. The overdensity is real, but no star is a permanent resident. As gas is swept into the crest it shocks, dust lanes form, and star formation ignites — which is why arms glow blue with young O and B stars while the old disk shows only a smooth, weak spiral. Made quantitative, one linearizes the coupled fluid (or collisionless Boltzmann) and Poisson equations about an axisymmetric equilibrium, and in the tightly-wound WKB limit this collapses to the Lin–Shu dispersion relation, tying the pattern frequency, the epicyclic frequency $\kappa$ , the surface density $\Sigma_0$ , and the velocity dispersion together. It singles out three radii — the inner and outer Lindblad resonances and corotation — that bound and drain the wave, while Toomre's $Q$ decides whether the disk is cool enough to carry such waves at all.

Key takeaways

A spiral density wave is a quasi-stationary crest in a disk's surface density and potential that rotates rigidly at a single pattern speed $\Omega_p$ , distinct from the differential rotation $\Omega(R)$ of the stars and gas that stream through it.
The Lin–Shu hypothesis kills the winding problem: a crest is a locus of constant phase, not of fixed matter, so a constant $\Omega_p$ never coils up — even though material at $R=8\ \mathrm{kpc}$ would wind to pitch angle $i<1°$ within a few Gyr.
Linearizing continuity, Euler, and Poisson in the WKB limit gives $(\omega-m\Omega)^2=\kappa^2-2\pi G\Sigma_0|k|+k^2c_s^2$ , a tug of war between rotation, self-gravity, and pressure.
The Toomre parameter $Q=\kappa c_s/(\pi G\Sigma_0)$ (fluid) or $\kappa\sigma_R/(3.36\,G\Sigma_0)$ (stars) sets local stability; the solar neighbourhood sits at $Q\approx1.9$ , cool enough to carry spirals yet stable against fragmentation.
Three resonances organize where waves can live: corotation ( $\Omega_p=\Omega$ ) and the inner and outer Lindblad resonances ( $\Omega_p=\Omega\mp\kappa/m$ ), which reflect and absorb wave energy.
A single steady eigenmode is idealized: real disks are powerful swing amplifiers and much structure is transient and recurrent — but density-wave theory remains the organizing language for spiral kinematics.
Key idea: A spiral arm is a wave in the disk's density and gravitational potential rotating at a single pattern speed $\Omega_p$ , not a fixed set of stars; the winding problem vanishes because a wave crest need not corotate with matter, and the local dynamics is governed by the WKB dispersion relation and the Lindblad and corotation resonances.

Where this lives today

Density-wave theory is the working grammar of modern galactic dynamics. The Tremaine–Weinberg method, which extracts $\Omega_p$ directly from the line-of-sight kinematics of any continuity-obeying tracer, is applied routinely to integral-field surveys like MaNGA and PHANGS–MUSE, pinning down corotation in hundreds of disks. In the Milky Way, ESA's Gaia mission turned the theory inward: the phase-space "snail shell" found in 2018 — a spiral in the $(z,v_z)$ plane of solar-neighbourhood stars — records the disk ringing after a perturbation, and its winding rate dates the last encounter with the Sagittarius dwarf, the winding argument run in reverse. ALMA and JWST resolve the predicted streaming across individual arms — dust lanes on the leading edge, a ridge of newborn O and B stars downstream, then an aging redder population — the color/age offset the Lin–Shu flow requires. Meanwhile N-body simulations show that live disks near $Q\sim1$ – $2$ are relentless swing amplifiers, generating recurrent transient spirals and driving radial migration across kpc — a secular reshaping in which the density wave is both cause and ruler.

The physical picture

Picture a two-lane roundabout where a slowly-rotating chevron of orange cones marks the "arm." Cars — stars and gas clouds — circulate faster than the cones inside corotation, bunch up as they climb into the cones' gravitational dip, then thin out as they escape. The bunching is a genuine overdensity, but no car stays in the jam: the pattern persists while matter pours through it. A spiral arm is a standing feature in the flow, not a rigid bar of stars welded to the disk — as the first figure makes concrete, marking the corotation circle and the timescale that makes the winding problem quantitative.

Figure

Stars stream through a rigidly rotating two-arm pattern. Inside the corotation circle (dashed, where Ω = Ωₚ) material overtakes the arm; outside it lags. The crest is a locus of constant phase, not fixed matter, so it never winds — even though a material spoke at R = 8 kpc would coil to pitch angle i < 1° within a few Gyr, since one galactic rotation takes ≈ 2.2×10⁸ yr

The winding problem forces the wave picture on us. A material spoke shears because each radius carries its own $\Omega(R)$ , and its pitch angle collapses as $\cot i = R\,|\mathrm{d}\Omega/\mathrm{d}R|\,t \approx \Omega\, t$ for a flat curve — reaching $i<1°$ within a few Gyr, far tighter than the observed $10°$ – $30°$ (Derivation 1). A wave escapes this entirely: its crest satisfies $m\phi - m\Omega_p t + \Phi(R)=\text{const}$ with $\Omega_p$ fixed, so the pattern turns rigidly while the material slips through it. Underneath the streaming sits the microscopic engine. A star on a nearly circular orbit oscillates radially at the epicyclic frequency $\kappa$ while it circulates at $\Omega$ ; in the rotating frame it meets the $m$ -armed wave once every $2\pi/[m(\Omega-\Omega_p)]$ . When that beat resonates with the epicycle — one oscillation per $m$ arm-crossings — the wave pumps angular momentum into or out of the orbit. Those are the Lindblad resonances, which with corotation bound where a wave can propagate. The second figure is the working diagram of the subject: $\Omega$ and $\Omega\pm\kappa/2$ against radius, cut by a horizontal pattern-speed line whose three intersections are the ILR, corotation, and OLR.

Figure

The resonance diagram for a flat rotation curve (v_c = 220 km/s, κ = √2 Ω). A two-arm pattern at Ωₚ = 20 km/s/kpc (horizontal line) meets Ω−κ/2 at the inner Lindblad resonance R ≈ 3.2 kpc, Ω at corotation R = 11 kpc, and Ω+κ/2 at the outer Lindblad resonance R ≈ 18.8 kpc. Waves live between the Lindblad resonances; the ILR sits at 0.293 R_CR and the OLR at 1.707 R_CR

Finally, stability is a tug of war that decides whether such waves exist at all. Self-gravity ( $G\Sigma_0$ ) wants to clump the disk; random motion ( $c_s$ or $\sigma_R$ ) resists small-scale collapse and shear ( $\kappa$ ) resists large-scale collapse. The Toomre parameter $Q$ names the local winner, and disks self-regulate near $Q\sim1$ – $2$ by a feedback loop — cool enough to amplify spiral waves, warm enough to avoid fragmentation — which is why nearly every disk lives on the same knife-edge.

The core equations

The pattern speed carries the whole idea: the wave's angular phase velocity is a single constant at every radius — what a rigidly rotating pattern means mathematically, and the assumption that dissolves the winding problem.

\Omega_p = \frac{\omega}{m} = \text{constant (independent of } R) \qquad (1)

where: $\Omega_p$ — angular pattern speed of the spiral (rigid-body rotation of the wave); $\omega$ — perturbation frequency of the $\exp[i(m\phi-\omega t)]$ mode; $m$ — azimuthal wavenumber, the number of arms; $R$ — galactocentric radius. Reading it: unlike $\Omega(R)$ , which falls outward, $\Omega_p$ is one number for the whole pattern. The epicyclic frequency is the disk's radial spring constant: nudge a circular orbit inward or outward and it oscillates in radius at $\kappa$ . This frequency, not $\Omega$ , is what the wave resonates against, so it appears in every relation below.

\kappa^2(R) = R\frac{\mathrm{d}\Omega^2}{\mathrm{d}R} + 4\Omega^2 = 2\Omega\left(2\Omega + R\frac{\mathrm{d}\Omega}{\mathrm{d}R}\right) \qquad (2)

where: $\kappa$ — epicyclic (radial oscillation) frequency of a nearly circular orbit; $\Omega(R)$ — circular angular velocity, with $\Omega^2 = (1/R)\,\mathrm{d}\Phi_0/\mathrm{d}R$ . Reading it: flat curve $\kappa=\sqrt{2}\,\Omega$ ; Keplerian $\kappa=\Omega$ ; in general $\Omega\le\kappa\le2\Omega$ . The Lin–Shu dispersion relation is the heart of the theory. Linearize the razor-thin fluid disk plus Poisson's equation, insert a tightly-wound plane wave, and demand a nontrivial solution: rotation ( $\kappa^2$ ) and pressure ( $k^2c_s^2$ ) restore the disk while self-gravity ( $-2\pi G\Sigma_0|k|$ ) destabilizes it, and when the right-hand side goes negative the disk is locally unstable.

(\omega - m\Omega)^2 = \kappa^2 - 2\pi G \Sigma_0 |k| + k^2 c_s^2 \qquad (3)

where: $k$ — radial wavenumber of the tightly-wound wave ( $|kR|\gg1$ ); $\Sigma_0$ — unperturbed surface mass density; $c_s$ — sound speed (effective velocity dispersion) of the gaseous disk; $G$ — Newton's constant. Reading it: the left side is the squared frequency in the rotating frame, the right the restoring spring less the self-gravitational pull. For a stellar disk the pressure term becomes a reduction factor $\mathcal{F}$ integrating the epicyclic response over a Gaussian velocity distribution: it softens self-gravity because hot orbits smear out over the wavelength, weakening their pull.

(\omega - m\Omega)^2 = \kappa^2 - 2\pi G \Sigma_0 |k|\,\mathcal{F}(s,\chi) \qquad (4)

where: $\mathcal{F}$ — Kalnajs–Lin–Shu reduction factor (a Landau-type integral); $s=(\omega-m\Omega)/\kappa=\nu$ — dimensionless Doppler-shifted frequency; $\chi=k^2\sigma_R^2/\kappa^2$ — dimensionless squared wavenumber measuring velocity-dispersion smearing; $\sigma_R$ — radial velocity dispersion of the stars. Reading it: as $\sigma_R\to0$ , $\mathcal{F}\to1$ and this reduces to the cold fluid relation. The Toomre parameter compresses local stability into one dimensionless number, emerging when we demand that no axisymmetric wavenumber has $\omega^2<0$ : above unity the disk is stable to all radial ripples, below unity it fragments.

Q \equiv \frac{\kappa\,c_s}{\pi G \Sigma_0}\;\;(\text{fluid}),\qquad Q \equiv \frac{\kappa\,\sigma_R}{3.36\,G\Sigma_0}\;\;(\text{stars}) \qquad (5)

where: $Q$ — Toomre stability parameter; $Q>1$ gives local axisymmetric stability, $Q<1$ instability; the constant $3.36$ comes from the Maxwellian stellar reduction factor (versus $\pi\approx3.14$ for a fluid). Reading it: cooler ( $c_s$ small) or denser ( $\Sigma_0$ large) disks have smaller $Q$ and collapse more readily. The resonance conditions locate the three organizing radii: corotation, where matter keeps pace with the pattern, and the Lindblad resonances, where a star completes one radial epicycle per $m$ arm encounters.

\nu = \frac{m(\Omega_p - \Omega)}{\kappa}:\quad \nu=0\;(\text{CR}),\qquad \nu=\pm1\;(\text{ILR / OLR}) \qquad (6)

where: CR — corotation resonance, $\Omega_p=\Omega$ ; ILR / OLR — inner / outer Lindblad resonance, $\Omega_p=\Omega\mp\kappa/m$ . Reading it: for $m=2$ , ILR at $\Omega_p=\Omega-\kappa/2$ , OLR at $\Omega_p=\Omega+\kappa/2$ . The group velocity says how fast a wave packet — and therefore wave energy — moves radially, making the density wave a dynamical object with a lifetime and a propagation cycle rather than a frozen picture.

v_{g} = \frac{\partial \omega}{\partial k} = \frac{k c_s^2 - \pi G \Sigma_0\,\mathrm{sgn}(k)}{\omega - m\Omega} \qquad (7)

where: $v_g$ — radial group velocity of the wave packet. Reading it: short trailing waves carry energy to the ILR where it is absorbed; long waves reflect near corotation — the feedback loop a self-sustained mode needs. The critical wavelength is the scale self-gravity most wants to collapse, set purely by $\Sigma_0$ and $\kappa$ — the natural length of spiral and ring structure.

\lambda_{\rm crit} = \frac{4\pi^2 G \Sigma_0}{\kappa^2},\qquad k_{\rm crit}=\frac{\kappa^2}{2\pi G\Sigma_0} \qquad (8)

where: $\lambda_{\rm crit}$ — longest axisymmetric wavelength destabilized by self-gravity in a cold disk (the boundary between rotation-stabilized long waves and pressure-needing short waves); $k_{\rm crit}$ — the corresponding wavenumber. Reading it: near the Sun a few kpc, matching the observed inter-arm spacing.

Derivation AThe winding problem and its resolution

The goal is to show, with numbers, that a spiral made of fixed stars is doomed while a wave is not. We follow a radial spoke of tracer stars laid across a differentially rotating disk and watch its pitch angle collapse under shear, then replace the material phase by a wave phase with a constant pattern speed and show that the crest rotates rigidly and never coils — the quantitative contrast is the entire motivation for density-wave theory.

Step 1. At $t=0$ mark stars along a radial line, $\phi(R,0)=\phi_0$ ; each then circulates at its own local speed, so $\phi(R,t)=\phi_0+\Omega(R)\,t$ . Why: each guiding-centre orbit simply advances at $\Omega(R)$ , with no force locking neighbouring radii together, so shear is unavoidable.

Step 2. The pitch angle obeys $\cot i = R\,|\mathrm{d}\phi/\mathrm{d}R| = R\,|\mathrm{d}\Omega/\mathrm{d}R|\,t$ . For a flat curve $\Omega=v_c/R$ , so $R\,|\mathrm{d}\Omega/\mathrm{d}R|=\Omega$ and $\cot i \approx \Omega\,t$ . Why: shear drives the feature monotonically from radial toward azimuthal — ever more tightly wound.

Step 3. Put in numbers: $\Omega\approx28\ \mathrm{rad\,Gyr^{-1}}$ (galactic year $\approx2.2\times10^8$ yr). After $1$ Gyr, $i\approx2°$ ; after $5$ Gyr, $i\approx0.4°$ — versus observed $i\sim10°$ – $30°$ . Why: the predicted material spiral is ten to a hundred times too tightly wound, so the winding problem is quantitative, not a quibble.

Step 4. Replace the material phase by a wave phase $m\phi - m\Omega_p t + \Phi(R)=\text{const}$ with $\Omega_p$ constant. The crest is now a locus of constant phase whose shape is frozen; the figure rotates rigidly at $\Omega_p$ and never winds, however wildly $\Omega(R)$ varies. Why: decoupling the pattern's rotation from the material's is exactly what a wave does — the crest need not be the same stars turn to turn.

Go deeper · Derivation B · 1 minThe Lin-Shu fluid dispersion relation

Derivation BThe Lin-Shu fluid dispersion relation

The goal is to derive equation (3) from first principles for a razor-thin gaseous disk, exposing where each of its three competing terms — rotation, self-gravity, pressure — comes from. We perturb the equilibrium with a tightly-wound plane wave, linearize continuity, the Euler equations, and Poisson's equation, then eliminate the velocity and potential perturbations to leave a single condition on the frequency. The WKB assumption is what makes the algebra local and the Poisson kernel simple.

Step 1. Start from the axisymmetric equilibrium $\Sigma=\Sigma_0(R)$ , $\mathbf{v}=(0,R\Omega)$ , and impose perturbations $\Sigma_1,\mathbf{v}_1,\Phi_1\propto\exp[i(kR+m\phi-\omega t)]$ with $|kR|\gg m$ . Why: tight winding lets the wave look locally like a plane wave in $R$ , so radial derivatives bring down factors of $ik$ and the problem becomes algebraic.

Step 2. Linearized continuity reads $-i(\omega-m\Omega)\Sigma_1 + ik\Sigma_0 v_{1R}=0$ , tying the density response to the radial velocity via the Doppler-shifted frequency $\omega-m\Omega$ . Why: this is conservation of mass carried by the mean flow — the crest is built and destroyed as gas streams through it.

Step 3. The linearized Euler equations fold Coriolis coupling $2\Omega v_{1\phi}$ and shear into the epicyclic frequency $\kappa^2$ , forced by the pressure-plus-gravity gradient $-ik(\Phi_1 + c_s^2\Sigma_1/\Sigma_0)$ , giving a response $\propto[\kappa^2-(\omega-m\Omega)^2]^{-1}$ . Why: the disk's natural radial oscillation is the epicycle, so forcing is resonantly amplified near $(\omega-m\Omega)^2=\kappa^2$ — the origin of the $\kappa^2$ term.

Step 4. Close with the WKB Poisson solution for a thin sheet, $\Phi_1=-2\pi G\Sigma_1/|k|$ . Why: a razor-thin sheet's self-potential falls off as $1/|k|$ in Fourier space, which makes self-gravity enter linearly in $|k|$ rather than as $k^2$ .

Step 5. Substitute back, demand a nontrivial $\Sigma_1\neq0$ solution, and collect: $(\omega-m\Omega)^2=\kappa^2-2\pi G\Sigma_0|k|+k^2c_s^2$ . As $G\Sigma_0,c_s\to0$ this reduces to free epicyclic oscillation $(\omega-m\Omega)^2=\kappa^2$ . Why: linking frequency to wavenumber makes it a dispersion relation, and it reproduces Binney and Tremaine's fluid result exactly.

Go deeper · Derivation C · 1 minThe Toomre criterion for axisymmetric stability

Derivation CThe Toomre criterion for axisymmetric stability

The goal is to find the sharp condition under which a rotating disk resists all axisymmetric ( $m=0$ ) ripples, packaged into the single parameter $Q$ . We set $m=0$ in the fluid dispersion relation, treat $\omega^2$ as a function of wavenumber, minimize it, and demand that even its lowest value stay non-negative — a clean threshold separating stable disks from ones that fragment, and the reason every observed disk sits close to the boundary.

Step 1. Set $m=0$ , leaving $\omega^2=\kappa^2-2\pi G\Sigma_0|k|+k^2c_s^2$ — an upward parabola in $|k|$ : rotation dominates at long wavelengths, pressure at short ones, with self-gravity dipping the curve between. Why: instability is only possible where the self-gravity dip drags $\omega^2$ below zero, so the danger lives at intermediate wavelengths.

Step 2. Minimize over $|k|$ : $\mathrm{d}\omega^2/\mathrm{d}|k|=0$ gives $|k|_{\min}=\pi G\Sigma_0/c_s^2$ and $\omega^2_{\min}=\kappa^2-(\pi G\Sigma_0)^2/c_s^2$ . Why: if the disk survives its single most unstable wavelength it survives all of them, so stability reduces to the sign of one number.

Step 3. Demand $\omega^2_{\min}\ge0$ : $\kappa c_s\ge\pi G\Sigma_0$ , i.e. $Q\equiv\kappa c_s/(\pi G\Sigma_0)\ge1$ . Why: this is the Toomre criterion — stable if and only if $Q\ge1$ , with the marginal disk trembling at $|k|_{\min}=\kappa/c_s$ .

Step 4. For a stellar disk the Maxwellian reduction factor replaces $\pi$ by $3.36$ , giving $Q=\kappa\sigma_R/(3.36\,G\Sigma_0)$ . Why: stars are collisionless, so the effective pressure is set by $\sigma_R$ and the different constant reflects the Gaussian, not adiabatic-gas, response.

Limiting cases

The dispersion relation and resonance structure simplify in instructive regimes, each a sanity check on the full theory:

Case	Result	Meaning
Cold fluid disk ( $c_s\to0$ )	$(\omega - m\Omega)^2 = \kappa^2 - 2\pi G \Sigma_0 \vert k\vert$	Pressure vanishes, so short wavelengths are no longer stabilized: unstable for $\vert k\vert > \kappa^2/(2\pi G\Sigma_0)$ , i.e. below $\lambda_{\rm crit}$ . Rotation still stabilizes the long waves; formally $Q=0$ .
Marginal Toomre stability ( $Q=1$ )	$\vert k\vert_{\min}=\kappa/c_s,\;\; \lambda=\tfrac12\lambda_{\rm crit}=2\pi^2 G\Sigma_0/\kappa^2$	Neutral point: one wavelength is marginally stable and all others stable. Disks near $Q=1$ are the most effective swing amplifiers.
Flat rotation curve	$\kappa=\sqrt{2}\,\Omega$	The Milky Way's regime: $\text{ILR}$ at $0.293\,R_{\rm CR}$ and $\text{OLR}$ at $1.707\,R_{\rm CR}$ for $m=2$ , independent of $\Omega_p$ .
Corotation resonance	$\Omega_p = \Omega(R_{\rm CR}),\quad \nu = 0$	Matter and pattern move together; separates leading from trailing propagation and is where swing amplification and corotation torques peak.
Lindblad resonances	$\Omega_p = \Omega(R_{\rm L}) \pm \kappa(R_{\rm L})/m,\quad \nu = \pm1$	Inner ( $-$ ) and outer ( $+$ ). Short trailing waves are absorbed (Landau-damped) at the ILR; wave propagation is confined between the resonances.

Common trapThe winding fallacy, and its two quieter cousins

The headline mistake is to treat a spiral arm as a fixed collection of stars — the winding fallacy, which predicts arms ten to a hundred times more tightly wound than observed. An arm is a wave crest matter flows through. Two subtler errors also trip up careful students. First, do not read the cold-disk limit as "long wavelengths collapse": rotation stabilizes long axisymmetric waves, and it is the short wavelengths (below $\lambda_{\rm crit}$ ) that need pressure to survive — Jeans's pressureless intuition is reversed by rotation. Second, do not assume a single steady eigenmode explains every galaxy: real disks are powerful swing amplifiers, so much structure is transient, and the WKB relation is only valid at small pitch angle while real arms ( $i\sim10°$ – $30°$ ) need non-WKB calculations.

Worked examples

1 · Is the solar neighbourhood stable? (fully worked)

Take the disk near the Sun: circular speed $v_c=220\ \mathrm{km\,s^{-1}}$ at $R=8\ \mathrm{kpc}$ , stellar radial velocity dispersion $\sigma_R=35\ \mathrm{km\,s^{-1}}$ , and disk surface density $\Sigma_0=50\ M_\odot\,\mathrm{pc^{-2}}$ . Compute $\Omega$ , $\kappa$ , and the stellar Toomre $Q$ , and interpret.

Step 1. The angular speed is $\Omega=v_c/R=220/8=27.5\ \mathrm{km\,s^{-1}\,kpc^{-1}}=8.9\times10^{-16}\ \mathrm{s^{-1}}$ , a rotation period $2\pi/\Omega\approx\mathbf{2.2\times10^{8}\ yr}$ — one galactic year.

Step 2. For a flat rotation curve $\kappa=\sqrt{2}\,\Omega=\mathbf{38.9\ km\,s^{-1}\,kpc^{-1}}$ : the radial spring is stiffer than the rotation, so epicyclic oscillations run faster than the orbit itself.

Step 3. Form $Q=\kappa\sigma_R/(3.36\,G\Sigma_0)$ . In SI ( $\Sigma_0=0.104\ \mathrm{kg\,m^{-2}}$ , $\kappa=1.26\times10^{-15}\ \mathrm{s^{-1}}$ , $\sigma_R=3.5\times10^4\ \mathrm{m\,s^{-1}}$ ) the numerator $\kappa\sigma_R=4.4\times10^{-11}$ and denominator $3.36\,G\Sigma_0=2.3\times10^{-11}$ , giving $Q\approx\mathbf{1.9}$ .

Step 4. Interpret: $Q>1$ , so the solar neighbourhood is stable to axisymmetric collapse, but $Q$ sits close to the marginal value — exactly where a disk is cool enough to support and swing-amplify spiral waves. The Milky Way disk near the Sun is marginally stable ( $Q\approx1.9$ ), the self-regulated state that makes spiral structure both possible and inevitable.

2 · Where are the resonances? (fully worked)

A two-armed ( $m=2$ ) spiral pattern turns at $\Omega_p=20\ \mathrm{km\,s^{-1}\,kpc^{-1}}$ in a flat-rotation-curve galaxy with $v_c=220\ \mathrm{km\,s^{-1}}$ . Locate corotation and the two Lindblad resonances.

Step 1. Corotation is where $\Omega=\Omega_p$ : $v_c/R_{\rm CR}=\Omega_p$ , so $R_{\rm CR}=v_c/\Omega_p=220/20=\mathbf{11\ kpc}$ .

Step 2. For a flat curve $\kappa=\sqrt{2}\,\Omega$ , so $\Omega\pm\kappa/2=\Omega(1\pm\sqrt{2}/2)$ , with factors $1-\sqrt{2}/2=0.293$ (inner) and $1+\sqrt{2}/2=1.707$ (outer), independent of $\Omega_p$ .

Step 3. The ILR has $\Omega_p=0.293\,\Omega$ , so $R_{\rm ILR}=0.293\times11=\mathbf{3.2\ kpc}$ ; the OLR has $\Omega_p=1.707\,\Omega$ , so $R_{\rm OLR}=1.707\times11=\mathbf{18.8\ kpc}$ .

Step 4. The wave is confined between $3.2$ and $18.8$ kpc, and short trailing waves reaching the ILR are Landau-absorbed there. The same disk's critical wavelength $\lambda_{\rm crit}=4\pi^2 G\Sigma_0/\kappa^2\approx5.6$ kpc matches the observed inter-arm spacing. A given $\Omega_p$ pins a whole resonance skeleton: ILR at $0.29\,R_{\rm CR}$ , OLR at $1.71\,R_{\rm CR}$ — which is why measuring $\Omega_p$ is the master key to a galaxy's spiral geometry.

Physics of the portrait

Spiral structure is where celestial mechanics, fluid dynamics, and stellar kinematics meet, and the person who welded them into a predictive theory was an applied mathematician.

Origins · 1 minC. C. Lin and how this was discovered

C. C. Lin

1916–2013 · MIT applied mathematician, co-author of the Lin–Shu density-wave theory

Chia-Chiao Lin was one of the twentieth century's great applied mathematicians, trained on the hydrodynamic stability of shear flows before turning that machinery on the Galaxy. Bertil Lindblad had argued since the 1920s that spiral arms must be tied to orbital resonances rather than made of fixed stars, but his kinematic picture never produced a self-consistent, self-gravitating wave. In 1964 Lin and his student Frank Shu supplied exactly that: the quasi-stationary spiral structure hypothesis — a grand-design spiral is a density wave rotating rigidly at a single pattern speed — closing the loop by requiring the wave's own gravity to sustain the orbits that make it. With Lin, Yuan and Shu (1969) they wrote down the tightly-wound WKB dispersion relation that still bears their names. The theory did not survive untouched — Toomre's stability criterion and the discovery of swing amplification showed that many spirals are transient rather than eternal modes — but Lin's reframing, arm as wave not object, is the permanent core and the language every subsequent theory speaks.

History

1927

— Lindblad introduces orbital resonances to explain spiral arms

1964

— Lin and Shu publish the quasi-stationary spiral density-wave hypothesis

1964

— Toomre derives the local axisymmetric stability parameter Q

1966

— Julian and Toomre identify swing amplification in sheared disks

1981

— Toomre reframes many spirals as recurrent transient instabilities

Going further

Texts. Binney & Tremaine, Galactic Dynamics, 2nd ed., Ch. 6 — the definitive treatment · Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., Ch. 25 · Mo, van den Bosch & White, Galaxy Formation and Evolution, §11 · original papers: Lin & Shu (1964), ApJ 140, 646; Lin, Yuan & Shu (1969), ApJ 155, 721; Toomre (1964), ApJ 139, 1217; Julian & Toomre (1966), ApJ 146, 810; Toomre (1981), in The Structure and Evolution of Normal Galaxies.

Provenance. Equations cross-checked against Binney & Tremaine, 2nd ed., Ch. 6: epicyclic frequency (§3.2), fluid dispersion relation eq. (6.55), Toomre criterion (§6.2), resonances (§3.3). A verifier should reconfirm: (1) $\kappa=\sqrt{2}\,\Omega$ (flat curve), $\kappa=\Omega$ (Keplerian); (2) $Q=\kappa c_s/(\pi G\Sigma_0)$ with stellar constant $3.36$ ; (3) $\lambda_{\rm crit}\approx5.6$ kpc and stellar $Q\approx1.9$ for $v_c=220\ \mathrm{km\,s^{-1}}$ , $R=8$ kpc, $\sigma_R=35\ \mathrm{km\,s^{-1}}$ , $\Sigma_0=50\,M_\odot\,\mathrm{pc^{-2}}$ ; (4) $R_{\rm ILR}=0.293\,R_{\rm CR}$ , $R_{\rm OLR}=1.707\,R_{\rm CR}$ ; (5) the cold-disk instability direction ( $|k|>\kappa^2/2\pi G\Sigma_0$ , i.e. short wavelengths). All figures are standard textbook values.

Prerequisites. Collisionless Boltzmann Equation · Orbits in Axisymmetric and Triaxial Potentials · Galactic Potential Models · Action-Angle Variables. Leads to. Jeans Equations · Gravitational N-Body Problem · Distribution Function and Phase Space.

Ask the tutor: Ask me to walk through how a disk sitting at $Q\approx1$ turns a faint leading ripple into a strong trailing spiral — the swing amplifier that makes real galaxies spiral.

Is a spiral arm actually more stars, or just brighter stars?

Both, but the second effect dominates what you see. The old stellar disk really is denser in the arm — but the overdensity is only 10–20%, a gentle ripple. What makes arms leap out in blue and UV is that gas shocks as it flows into the crest, triggering star formation; the young, hot, short-lived O and B stars are enormously luminous, so the arm shines far out of proportion to its mass contrast. An arm is a modest overdensity wearing a spectacular coat of newborn stars.

What is the difference between the pattern speed and the rotation speed?

The rotation speed $\Omega(R)$ is how fast the material at radius $R$ orbits, and it decreases outward. The pattern speed $\Omega_p$ is how fast the spiral figure rotates, a single constant. Inside corotation the material overtakes the arms, outside it lags, and only at corotation do they coincide. Keeping them distinct is the whole content of the density-wave idea; conflating them is the winding fallacy.

Why does the Toomre parameter use κ and not Ω?

Because the instability $Q$ guards against is radial collapse, and the disk's restoring force against radial displacement is the epicyclic spring, whose frequency is $\kappa$ , not $\Omega$ . A star nudged inward does not simply fall in; conservation of angular momentum swings it back out at frequency $\kappa$ . For a flat curve $\kappa=\sqrt2\,\Omega$ , but in general they are independent — a strongly shearing disk can have $\kappa$ well below $2\Omega$ . Using $\kappa$ correctly captures how rotation stabilizes long-wavelength collapse.

What exactly happens at a Lindblad resonance?

At a Lindblad resonance a star completes exactly one radial epicycle for every $m$ times it meets the arm, so it always encounters the arm at the same phase of its radial oscillation. The wave's weak gravitational tug, instead of averaging away, adds up coherently turn after turn — a mechanical resonance — driving a strong exchange of energy and angular momentum. Short trailing waves reaching the inner Lindblad resonance are Landau-absorbed there, draining the wave; the ILR and OLR act as the walls of the box the spiral lives in, and the ILR can bar a two-armed wave from the very center.

If most spirals are transient, is Lin–Shu theory wrong?

No — incomplete rather than wrong. The picture of a single eternal eigenmode was an idealization: N-body simulations show that live disks near $Q\sim1$ – $2$ are strong swing amplifiers that continually generate and dissolve spirals, so flocculent and multi-armed galaxies are best described as recurrent transients. But every transient is still a density wave locally — it obeys the dispersion relation, has resonances, and produces the streaming and color gradients Lin–Shu predicts. The framework, arm as wave governed by $\kappa$ , $\Sigma_0$ , $Q$ , and the resonances, survives intact; what changed is the lifetime of the modes, not their physics.

How do observers actually measure a pattern speed?

Two ways. The classical route uses the color/age gradient: gas overtakes the arm inside corotation and lags outside, so the offset between the dust lane and the downstream young-star ridge flips sign at corotation, giving $R_{\rm CR}$ and hence $\Omega_p=\Omega(R_{\rm CR})$ . The modern, model-independent route is the Tremaine–Weinberg method: integrate the line-of-sight velocity and surface brightness of a continuity-obeying tracer along lines parallel to the major axis, and the ratio of two integrals gives $\Omega_p$ directly.