Interplanetary dust

Interplanetary dust consists of small solid particles orbiting within a planetary system and responding to gravity, radiation, plasma, and collisions.

A dust tail is an extended distribution of grains released from a localized moving source. Radiation pressure changes the grains' orbits; their release times, grain properties and survival probabilities determine the spatial profile.

For a single-stream dust tail with injection rate $\dot{N}$, monotone angular distance $s(t)$, and survival probability $P(t)$, the number per unit angle is $n(s)=\dot{N}P(t(s))/\dot{s}$. A vanishing angular drift creates an idealized density enhancement; velocity dispersion rounds it.

A zero of the angular drift in a cold single-grain-size dust tail produces a formally divergent number density per unit angle. The divergence is integrable when the exact local angle-to-age map is retained. A spread of launch velocities or radiation-pressure coefficients removes the singularity.

For unwrapped grain and parent azimuths, define a signed lag $\Delta \theta ={\theta }_d-{\theta }_p$ and a positive trailing distance $s=-\Delta \theta$. A grain released without a velocity kick from a circular parent orbit has $\Delta \theta =2\beta (\sin {\theta }_d-{\theta }_d)+O({\beta }^2{\theta }_d)$. Keeping track of this sign avoids mistaking trailing motion for leading motion.

The replacement $s=2\beta (\theta -\sin \theta )\mapsto 2\beta \theta$ neglects a bounded orbital ripple in accumulated lag. Substituting the resulting angle into the instantaneous periodic drift gives a convenient ${\csc }^2(s/(4\beta ))$ profile, but is not a uniform small-$\beta$ approximation. The consistent first-order profile retains the implicit angle-to-lag map; near release it gives $n(s)\propto s^{-2/3}$.

The radiation-pressure coefficient is the ratio of outward stellar radiation pressure to inward stellar gravity on a dust grain. The grain therefore feels effective gravitational parameter $(1-\beta )G{M}_{\ast }$.

If release imparts no velocity kick, a grain's initial Kepler orbit under reduced stellar attraction has $q=r_b$, $Q=r_b/(1-2\beta )$, and $e=\beta /(1-\beta )$ for $0\le \beta <1/2$. Its pericentre distance remains at the dust birth ring, while its apocentre distance grows with the radiation-pressure coefficient.

The radiation-pressure efficiency is the radiation-force cross-section of a grain divided by its geometric cross-section. It is of order unity for grains larger than the illuminating wavelength and depends strongly on size and composition in the small-particle regime.

A grain released from a circular parent orbit without a relative velocity impulse is unbound when radiation pressure reduces the effective stellar attraction sufficiently. For $\beta >1$, the effective inverse-square force is repulsive.

The zero-kick circular-parent threshold is $\beta =1/2$: below it grains are initially bound, at it their conservative release orbit is parabolic, and above it they are initially unbound. Poynting–Robertson drag can modify a marginal orbit, so the threshold is a statement about the instantaneous release orbit, not an exact capture criterion with dissipation included.

The Finson–Probstein model follows cometary dust after release with negligible ejection speed under stellar gravity reduced by radiation pressure. A grid of synchrones and syndynes organizes the resulting dust-tail morphology.

A synchrone is the locus, at one observation time, of grains released at the same time with different radiation-pressure coefficients.

A syndyne is the locus, at one observation time, of grains with the same radiation-pressure coefficient released at different times.

For an axisymmetric steady dust outflow with surface density $\Sigma$ and radial speed $v_r$, number or mass conservation gives $2\pi r\Sigma v_r=\text{constant}$.

Poynting–Robertson drag is the tangential component of stellar radiation force caused by aberration in an orbiting dust grain's frame. It removes orbital energy and angular momentum and drives the grain inward.

For grains of fixed diameter $D$ released instantaneously at $r_0$, the drift law $\dot{r}=-\gamma /(rD)$ integrates to the displayed size-radius map. The time to radius $r<r_0$ is $t_r(D)=D(r_0^2-r^2)/(2\gamma )$. Grains therefore arrive in increasing diameter order. This model of Poynting–Robertson drag assumes nearly circular Kepler orbits, fixed positive $\gamma$, and no grain loss or diameter changes before arrival. At a positive time, the map is monotone in $D$ and allows the initial power-law size distribution to be transformed into a spatial distribution or an arrival-time distribution.

For initial dust mass $m_d$ in a ring at $r_0$ with a fixed-diameter power-law size distribution of index $\alpha <4$, burst transport under inverse-size radial drag maps radius to diameter $D_r=2\gamma t/(r_0^2-r^2)$. Set $T_r=t_r(D_{\max })$ and $\eta =D_{\min }/D_{\max }$. The surface density is the azimuthal average: $2\pi r\Sigma dr=(\pi \rho K/6)D_r^{3-\alpha }d{D}_r$. Since $d{D}_r/dr=2r{D}_r/(r_0^2-r^2)$, mass normalization of a power-law size distribution gives the displayed expression. It is supported where $\eta \le t/T_r\le 1$. It applies at $t>0$ before removal at the radius in question; the initial state is a delta-function ring. A local angular density needs additional phase information.

Let the initial dust mass be $m_d$, let its differential number distribution be $K{D}^{-\alpha }$ with $\alpha <4$ between $D_{\min }$ and $D_{\max }$, and put $\eta =D_{\min }/D_{\max }$. Under burst transport under inverse-size radial drag, complete absorption at a fixed radius occurs at $t=TD/D_{\max }$, with $T=t_r(D_{\max })$. The arriving mass is proportional to $D^{3-\alpha }dD$. Substitute $D=D_{\max }t/T$ and $dD/dt=D_{\max }/T$ and use mass normalization of a power-law size distribution to prove the displayed rate for $\eta T\le t\le T$, with zero rate elsewhere. Its integral is exactly $m_d$. Resonant delays, scattering losses, a finite release duration or grain destruction change this ideal pulse.

Consider a circular circumplanetary orbit well inside the Hill sphere, with dust speed $u^2=G{M}_p/a$ and unit plane normal $\hat{\mathbf{k}}$. During one dust orbital period, take the stellar direction $\hat{\mathbf{R}}$ and flux at the planet's stellar radius $a_p$ as constant. The velocity-dependent Poynting–Robertson drag has relative workThe constant terms have zero work over the closed circular orbit. Since $⟨u_i{u}_j⟩=(u^2/2)({\delta }_{ij}-k_i{k}_j)$,The coefficient is three for coplanar stellar and dust orbits; it is two for a dust plane normal to the instantaneous stellar direction. Averaging over a circular planetary year at fixed dust orbital inclination $I$ gives $3-{\sin }^{2}I/2$. Conservative perturbations must remain small enough for this circular-orbit average.

For coplanar circular dust, constant $A=\beta G{M}_{\star }/c$, fixed stellar radius $a_p$ and initial planetocentric radius $a_0>R_p$, integrating $\dot{a}=-3Aa/a_p^2$ givesCompare the inspiral time under Poynting–Robertson drag for a circular circumstellar orbit, $t_{\mathrm{cs}}=(a_p^2-R_{\star }^2)/(4A)$. Planetary arrival is not automatically faster; in the point-star limit it is faster only for $a_0/R_p<e^{3/4}$. A different constant orientation coefficient replaces three.

With $A=\beta G{M}_{\star }/c$ constant, average the drag work $\dot{\epsilon }=-A(2{\dot{r}}^2+v_{\theta }^2)/r^2$ over a bound elliptic orbit. Using $dt=r^{2}df/h$ gives the integral $\pi (2+3e^2){\mu }^2/h^2$, where $\mu$ is the conservative central parameter. Combining with $\dot{a}=2a^2\dot{\epsilon }/\mu$ givesThis is a weak-drag secular average. It also holds when the conservative parameter is $(1-\beta )G{M}_{\star }$, provided a bound orbit exists; the conservative parameter cancels but the drag amplitude remains $A$. For a circular Kepler orbit, $a^2(t)=a_0^2-4At$.

For a blackbody grain with geometric cross-section $\sigma$, mass $m$ and stellar-frame velocity $\mathbf{v}$, define the radiation-pressure coefficient $\beta =L_{\star }\sigma /(4\pi cG{M}_{\star }m)$. To first order in $v/c$, steady absorption and isotropic rest-frame reradiation giveThe relative incoming flux provides $1-\dot{r}/c$; photon reradiation carries stellar-frame momentum proportional to $\mathbf{v}/c$. Subtracting static radiation pressure gives radial drag $-2m\beta G{M}_{\star }\dot{r}/(c{r}^2)$ and transverse drag $-m\beta G{M}_{\star }{v}_{\theta }/(c{r}^2)$. Non-blackbody grains require their actual radiation-pressure efficiency.

For the Poynting–Robertson drag invariant $C$ and $A=\beta G{M}_{\star }/c$, the orbit-averaged time to the point-star limit isThis tends to $a_0^2/(4A)$ for a circular orbit and to $4Q_0^{1/2}{q}_0^{3/2}/(5A)$ for a highly eccentric orbit. Stellar radius, sublimation, collisions, and failure of orbital averaging can terminate the evolution earlier.

The orbit-averaged Poynting–Robertson drag equations preserve $C=a(1-e^2)e^{-4/5}$ for $0<e<1$. This relates shrinking semi-major axis to decreasing orbital eccentricity; a circular orbit is treated by its separate limiting solution.

Stellar-wind drag is the tangential force exerted by a star's outflow on an orbiting grain. Like Poynting–Robertson drag, it removes orbital energy and angular momentum.

During inward migration, dust can be captured into an exterior mean-motion resonance with a planet. A phase-shifted conjunction then lets planetary torque replace angular momentum lost to Poynting–Robertson drag.

For an exterior $(p+1):p$ mean-motion resonance, use mean longitude in $\varphi =(p+1)\lambda -p{\lambda }_p-\varpi$, $z=e\exp (i\varphi )$, $x=(a-a_r)/a_r$ and $B=3p{n}_p/2$. Given $\dot{e}=-C\sin \varphi -Ae$ and $\dot{\varpi }=(C/e)\cos \varphi$, the leading Keplerian detuning gives $\dot{\varphi }=-Bx-(C/e)\cos \varphi$. Differentiating $z$ and using $e^{i\varphi }(\sin \varphi +i\cos \varphi )=i$ proves the displayed equation. It retains resonant forcing and eccentricity damping without the coordinate singularity at $e=0$. True longitude adds a fast orbital ripple and does not yield this autonomous averaged equation directly.

For positive locally constant $A,C$, supplement damped first-order resonant complex eccentricity by $\dot{a}/a=-2(p+1)Ce\sin \varphi -4A/5$. Balancing the two dissipative rates gives $e_f^2=2/[5(p+1)]$, $\sin {\varphi }_f=-A{e}_f/C$ and $B{x}_f=-(C/e_f)\cos {\varphi }_f$. Two fixed points exist when $A<C\sqrt{5(p+1)/2}$. The negative-cosine branch has $x_f>0$ and is stable in this leading model: a phase increase raises the positive resonant torque, raises semi-major axis, lowers mean motion, and reverses the phase drift. The positive-cosine branch has the opposite feedback.

For a coupled verification put $z=u+iv$. The characteristic polynomial at either fixed point is ${\lambda }^3+2A{\lambda }^2+[A^2+B^2{x}_f^2-2B(p+1)C{u}_f]\lambda -4AB(p+1)C{u}_f$. For $u_f<0$ its positive coefficients satisfy the cubic Routh-Hurwitz stability criterion, since the remaining inequality reduces to $2A(A^2+B^2{x}_f^2)>0$. For $u_f>0$ its negative constant term forces a positive real root. At the equality threshold the equilibria merge and an eigenvalue is zero. This is local stability for the frozen smooth-coefficient approximation, not a conclusion for arbitrary rapidly varying dissipation laws.

For resonant trapping of dust in an exterior $(p+q):p$ mean-motion resonance, let Poynting–Robertson drag have parameter $A$. Holding semi-major axis $a$ fixed and retaining secular rates through second order in orbital eccentricity givesFor $5p>q$, squared orbital eccentricity approaches this equilibrium exponentially in the truncated model. If the predicted value is not small, higher-order terms are needed for quantitative dynamics. Finite planetary torque also imposes a minimum eccentricity for balance, which alone does not guarantee capture during migration.

At pericentre, mean longitude equals longitude of periapsis, so resonant argument $\varphi =(p+q)\lambda -p{\lambda }_p-q\varpi$ becomes $\varphi =p(\varpi -{\lambda }_p)$. The pericentre angle in the planet's rotating frame is $\varphi /p$ modulo $2\pi /p$. For coprime $p,q$, there are $p$ distinct pericentre directions before the trajectory closes. Resonant-argument libration restricts these directions and supplies resonance protection.