Hill Sphere

The Hill sphere (or Hill radius, r_H; also known in French astronomical literature as the Roche sphere) is the roughly spherical region of space surrounding a secondary astronomical body of mass m (such as a planet orbiting a star, or a moon orbiting a planet) within which the secondary body's own gravity dominates the orbital motion of a tertiary test particle (such as a natural satellite or spacecraft) over the differential tidal pull of the much more massive primary body of mass M. Formulated independently by American mathematical astronomer George William Hill (1878) in his lunar theory and by French mathematician Édouard Roche, the Hill sphere corresponds dynamically to the equipotential surface passing through the inner and outer collinear Lagrange points (L₁ and L₂) of the circular restricted three-body problem.

For a secondary body of mass m (≪ M) orbiting a primary body of mass M with semi-major axis a and orbital eccentricity e, the characteristic radius of the Hill sphere at pericenter (closest approach, r_min = a(1 − e)) is:

r_H ≈ a (1 − e) (m / [3(M + m)])^1/3 ≈ a (1 − e) (m / [3M])^1/3

To remain bound in a stable long-term orbit around a planet, a satellite's orbital radius must lie well inside the planet's Hill sphere (typically within 0.33 to 0.50 r_H for prograde orbits and 0.65 to 0.70 r_H for retrograde orbits), while Simultaneously remaining outside the planet's Roche limit so that tidal forces do not tear the satellite apart.

Consider a primary star of mass M and a planet of mass m (with m ≪ M) revolving in circular orbits of separation a about their common center of mass at constant angular velocity Ω = √[G(M + m) / a³] ≈ √(GM / a³). In a co-rotating Cartesian frame centered on the planet along the star–planet line (the x-axis), a test particle at small displacement x ≪ a from the planet toward or away from the star experiences three radial accelerations:

1. Gravitational attraction toward the planet: g_planet = −Gm / x²
2. Differential gravitational attraction from the star (expanded to first order in x/a via the tidal quadrupole approximation):

g_star(a + x) − g_star(a) = −GM / (a + x)² + GM / a² ≈ + (2*GM* / a³) x = +2Ω² x

1. Differential centrifugal acceleration in the rotating frame:

g_cent(a + x) − g_cent(a) = Ω²(a + x) − Ω²a = +Ω² x

Summing the stellar tidal tension (+2Ω²x) and the centrifugal term (+Ω²x) yields the total outward tidal perturbing acceleration (+3Ω²x) along the star–planet line. The boundary of the Hill sphere—corresponding to the L₁ and L₂ saddle points of the Jacobi effective potential Φ_eff = −(GM/r₁ + Gm/r₂ + ½Ω²r²)—occurs where the planet's inward gravitational pull exactly balances the outward 3Ω²x tidal acceleration:

Gm / r_H² = 3 Ω² r_H = 3 (GM / a³) r_H

Solving for r_H yields Hill's fundamental scaling law:

r_H = a (m / [3M])^1/3

Because the three-dimensional equipotential surface passing through L₁ and L₂ is slightly elongated along the star–planet axis (x) and compressed along the orbital direction (y) and polar axis (z), the zero-velocity surface forms a lemon-shaped Hill lobe with principal axes in the ratio (x : y : z) = (1 : 2/3 : 0.638) (Murray & Dermott, 1999).

Because r_H scales linearly with semi-major axis a and only as the one-third power of planetary mass (m^1/3), distant outer ice giants possess vastly larger Hill spheres than close-in rocky planets or hot Jupiters:

Notice that Neptune—despite having only 18% of Saturn's mass and 5.4% of Jupiter's mass—possesses the largest Hill sphere of any planet in the Solar System (115 million km, or 0.77 AU) because it orbits at 30.1 AU where solar tidal gradients (∝ a⁻³) are exceedingly weak. Consequently, Neptune's outermost irregular retrograde moon, Neso, orbits at a mean distance of 49.3 million km (more than one-third of the Earth–Sun distance), taking 26.67 Earth years to complete a single revolution around Neptune.

Although the Hill radius r_H marks the zero-velocity boundary where the L₁ and L₂ gateways first open, numerical integrations and analytical perturbation studies (Hamilton & Burns, 1991; Holman & Wiegert, 1999; Nesvorný et al., 2003) demonstrate that test particles placed in circular orbits beyond ~0.5 r_H are rapidly stripped from the planet over tens to thousands of orbits. Furthermore, orbital stability inside the Hill sphere exhibits a strong prograde-versus-retrograde asymmetry governed by the Coriolis acceleration (−2 Ω** × v)** in the rotating frame:

• Prograde Satellites (Orbiting in the Same Sense as the Planet's Orbit Around the Star): In the frame co-rotating with the planet's orbital motion Ω, the Coriolis force on a prograde satellite points radially outward, adding to the centrifugal and stellar tidal forces and tending to destabilize the orbit. Moreover, prograde satellites at large radii encounter the evection resonance (where the satellite's apsidal precession period equals the planet's orbital period). Consequently, prograde satellites are stable over the 4.5-billion-year age of the Solar System only out to:

r_{crit,prograde} ≈ 0.33 to 0.49 r_H

• Retrograde Satellites (Orbiting Opposite to the Planet's Orbit Around the Star): For a retrograde satellite, the velocity vector v in the rotating frame is reversed, so the Coriolis acceleration (−2 Ω × v) points radially inward toward the planet, reinforcing planetary gravity against solar tides. As a result, retrograde orbits remain dynamically stable out to substantially larger radii:

r_{crit,retrograde} ≈ 0.65 to 0.70 r_H

This Coriolis stability asymmetry explains a striking observational feature of the giant planets: among the more than 200 irregular captured moons orbiting Jupiter, Saturn, Uranus, and Neptune at distances between 0.15 r_H and 0.50 r_H, over 80% occupy retrograde orbits, and all of the outermost irregular moons (such as Jupiter's Pasiphae and Carme groups and Neptune's Neso and Psamathe) are retrograde.

Four distinct gravitational boundary radii appear across celestial mechanics and astrophysics, each answering a different physical question:

1. Hill Sphere (r_H = a(m/[3M])^1/3): The tidal boundary for a small secondary mass (m ≪ M) within which a tertiary satellite remains gravitationally bound over complete orbits.
2. Roche Lobe (R_L, Eggleton 1983): The exact tear-drop-shaped equipotential surface passing through L₁ in a close binary star system of arbitrary mass ratio q = M₁/M₂ (where m is not necessarily small compared to M). When an evolving star expands past its Roche lobe, gas streams through the L₁ nozzle onto the companion star (Roche-lobe overflow, fueling Bondi/disk accretion, X-ray binaries, and Type Ia supernovae). In the limit q ≪ 1, the Roche lobe reduces to the Hill lobe.
3. Roche Limit (d_Roche ≈ 2.44 R_p (ρ_p/ρ_s)^1/3): The inner distance from a planet of radius R_p and density ρ_p inside which a fluid satellite of density ρ_s is pulled apart by tidal forces exceeding its own self-gravity. Thus, a natural moon can exist only in the annular zone where d_Roche < r_orbit < 0.5 r_H. For a hypothetical moon around an asteroid or around a close-in hot Jupiter where 0.5 r_H ≤ d_Roche, no stable moon can exist at all.
4. Laplace Sphere of Influence (r_SOI = a(m/M)^2/5): Used in astrodynamics for patched-conic hyperbolic flyby trajectories (such as Voyager or Cassini), defined as the radius where the ratio of perturbing acceleration to primary acceleration equals the ratio inside versus outside (scaling as (m/M)^2/5 rather than (m/M)^1/3). For Earth, r_SOI ≈ 925,000 km, slightly smaller than Earth's Hill radius (r_H ≈ 1,471,000 km).

In the core-accretion theory of planet formation within a protoplanetary disk (Kokubo & Ida, 1998), growing planetary embryos clear solid planetesimals from an annular feeding zone whose half-width is proportional to the embryo's Hill radius (Δa ≈ B · r_H, where B ≈ 5–10 Hill radii). As the embryo accretes surface mass Σ_s from this annulus, its mass m grows and its Hill radius r_H ∝ m^1/3 expands until it consumes all solid planetesimals within Δa, stalling at the oligarchic isolation mass:

M_iso = (8 / √3) (π B a² Σ_s)^3/2 / M_*^1/2

In the inner Solar System at 1 AU (inside the water snow line where Σ_s is small and a is small), M_iso is only ~0.05 to 0.10 M_⊕ (roughly the mass of Mars or Moon-to-Mars-sized embryos, requiring late giant impacts over 30–100 Myr to assemble Earth and Venus). Beyond the water snow line at ~2.7–5 AU, both a² and surface solid density Σ_s (boosted 3-to-4-fold by condensing water ice) jump simultaneously, raising M_iso to ~5 to 10 M_⊕—the critical core mass required to trigger runaway gas accretion and form gas giants like Jupiter and Saturn before the protoplanetary gas disk dissipates.

1. Hill, G. W. (1878). "Researches in the Lunar Theory." American Journal of Mathematics, 1(1), 5–26. https://doi.org/10.2307/2369430
2. Eggleton, P. P. (1983). "Approximations to the radii of Roche lobes." The Astrophysical Journal, 268, 368–369. https://doi.org/10.1086/160960
3. Hamilton, D. P., & Burns, J. A. (1991). "Orbital stability zones about asteroids." Icarus, 92(1), 118–131. https://doi.org/10.1016/0019-1035(91)90039-V90039-V)
4. Kokubo, E., & Ida, S. (1998). "Oligarchic Growth of Protoplanets." Icarus, 131(1), 171–178. https://doi.org/10.1006/icar.1997.5840
5. Nesvorný, D., Alvarellos, J. L. A., Dones, L., & Levison, H. F. (2003). "Orbital and Collisional Evolution of the Irregular Satellites." The Astronomical Journal, 126(1), 398–429. https://doi.org/10.1086/375461