Von Zeipel Theorem

The von Zeipel theorem is a foundational theorem of stellar structure and radiative transfer proved in 1924 by Swedish astronomer Edvard Hugo von Zeipel, which establishes two fundamental consequences for rotating stars: (1) the von Zeipel gravity-darkening law, which states that the local radiative energy flux F(θ) emerging from the surface of a uniformly rotating star in radiative equilibrium is directly proportional to the magnitude of its local effective surface gravity g_eff(θ) (F ∝ g_eff, so T_eff(θ) ∝ g_eff^1/4); and (2) the von Zeipel paradox, which proves that a uniformly rotating star in strict hydrostatic equilibrium cannot simultaneously maintain local radiative thermal equilibrium (∇ · F = ρ ε) throughout its interior—forcing all rotating radiative stars to develop slow, global Eddington–Sweet meridional circulation currents that mix chemical elements between the stellar core and surface (von Zeipel, 1924; Maeder, 2009).

Consider an axisymmetric star rotating at a uniform angular velocity Ω (a conservative centrifugal force field). In the corotating frame, the star is governed by the equation of hydrostatic equilibrium:

(1 / ρ) ∇P = −∇Φ_eff = g_eff

where P is total pressure, ρ is mass density, g_eff = −∇Φ_eff is the effective gravity vector (the vector sum of inward Newtonian gravitational acceleration −∇Φ_grav and outward centrifugal acceleration +½ Ω² ∇[r² sin²θ]), and Φ_eff(r, θ) is the total Roche equipotential:

Φ_eff(r, θ) = Φ_grav(r, θ) − ½ Ω² r² sin²θ

Taking the curl of the hydrostatic equation (∇ × [∇P / ρ] = −∇ × ∇Φ_eff = 0) yields ∇ρ × ∇P = 0, proving Poincaré's barotropic theorem: across a uniformly rotating star, surfaces of constant pressure (P), constant density (ρ), and—via the ideal gas and radiation equation of state P(ρ, T)—constant temperature (T) and Rosseland mean opacity (κ) coincide with surfaces of constant effective potential Φ_eff = constant (level surfaces / isobars).

Now apply the equation of radiative diffusion for the outward radiative energy flux F across an optically thick stellar envelope:

F = − [ 4 a c T³ / (3 κ ρ) ] ∇T

Because T, ρ, and κ are functions of the single scalar variable Φ_eff on every level surface, we can apply the chain rule ∇T = (dT / dΦ_eff) ∇Φ_eff = −(dT / dΦ_eff) g_eff:

F(θ) = − C(Φ_eff) g_eff(θ) ⇒ |F(θ)| ∝ g_eff(θ)

where C(Φ_eff) = [4*acT*³ / (3κρ)] (dT/dΦ_eff) is constant along the stellar surface (which is itself a single equipotential surface Φ_eff = Φ_surf). Applying the Stefan–Boltzmann law (F(θ) = σ_SB T_eff⁴(θ)) yields von Zeipel's gravity-darkening law:

T_eff(θ) ∝ [ g_eff(θ) ]^β with β = 0.25 (for a radiative envelope)

Why is the effect called gravity darkening (rather than centrifugal darkening)? In a rapidly rotating star (such as an O, B, or A main-sequence star rotating at 50%–95% of its Keplerian breakup velocity v_crit = √[GM / R_eq]), centrifugal force distorts the star into an oblate spheroid (Roche spheroid) whose equatorial radius R_eq is up to 1.5 times larger than its polar radius R_pole:

• At the Poles (θ = 0°, sin θ = 0): Centrifugal acceleration is zero and the surface lies closest to the stellar center of mass (R_pole is minimum). Therefore, effective surface gravity g_eff,pole = GM / R_pole² is maximum, equipotential surfaces are compressed (dT/dr is steep), and the poles are hot and luminous.
• At the Equator (θ = 90°, sin θ = 1): The surface bulges farthest from the core (R_eq is maximum) and outward centrifugal acceleration Ω²R_eq subtracts from inward gravity:

g_eff,eq = (GM / R_eq²) − Ω² R_eq. As Ω approaches the critical (breakup) rotation rate Ω_crit, g_eff,eq approaches zero, equipotentials spread apart, and the equatorial belt becomes cool and dark.

Radiative (von Zeipel, β = 0.25) vs. Convective (Lucy, β ≈ 0.08) Envelopes and the Espinosa Lara (2011) Baroclinic Correction

1. Convective Envelopes (Lucy's Law, 1967): In cooler stars (T_eff < 6,500 K, spectral types F, G, K, M) where energy is transported to the photosphere by turbulent convection rather than radiative diffusion, specific entropy S rather than radiative conductivity is constant along adiabatic streamlines. Leon B. Lucy (1967) showed that convective envelopes obey a weaker gravity-darkening exponent: β ≈ 0.08 (F ∝ g_eff^0.32).
2. 2D Non-Barotropic Radiative Models (Espinosa Lara & Rieutord, 2011): Because real rotating stars are slightly baroclinic (temperature varies on equipotential surfaces near the surface when Ω → Ω_crit), Francisco Espinosa Lara and Michel Rieutord (2011) derived the two-dimensional radiative flux solution without assuming barotropicity, showing that for rapid rotators (Ω/Ω_crit > 0.8), the effective exponent decreases from β = 0.25 down to β ≈ 0.14–0.21, preventing T_eff,eq from dropping to zero at breakup.

Beginning in 2001, long-baseline optical and near-infrared stellar interferometers—such as the CHARA Array on Mount Wilson (330-meter baseline, angular resolution ~0.5 milliarcseconds) and the VLTI (Very Large Telescope Interferometer) in Chile—resolved the oblate disks and polar hot spots of nearby rapid rotators (Monnier et al., 2007; Domiciano de Souza et al., 2003):

Furthermore, when a transiting exoplanet crosses an oblate, gravity-darkened host star along an inclined or polar chord (such as KOI-13 b / Kepler-13Ab or WASP-33 b), the planet sequentially occults the bright polar cap and the dim equatorial belt, imprinting an asymmetric W-shaped distortion on the photometric transit light curve that independently measures the 3D spin–orbit obliquity ψ alongside the Rossiter–McLaughlin effect (Barnes, 2009).

In the second half of his 1924 paper, von Zeipel calculated the divergence of the radiative flux ∇ · F inside a uniformly rotating barotropic star and set it equal to the local nuclear energy generation rate ρ ε:

∇ · F = − C(Φ_eff) ∇ · g_eff − (dC / dΦ_eff) g_eff² = 4π*Gρ* C − 2Ω² C − (dC / dΦ_eff) g_eff² = ρ ε

Because ρ, ε, and C are constant on an equipotential surface Φ_eff, whereas g_eff² varies with colatitude θ along that same surface, this equation can only be satisfied if dC/dΦ_eff = 0 and ε ∝ (1 − Ω² / [2π*Gρ*])—an unphysical condition never met by real nuclear fusion rates! This contradiction is the von Zeipel Paradox: a uniformly rotating star in radiative equilibrium cannot remain in local thermal balance (∇ · F ≠ ρ ε).

As Arthur S. Eddington (1925) and P. A. Sweet (1950) demonstrated, the slight latitudinal heating imbalance (∇ · F < ρ ε along the polar axis; ∇ · F > ρ ε near the equator) creates buoyancy forces that drive a global quadrupolar thermal convection current across the radiative zone—Eddington–Sweet meridional circulation—in which gas rises slowly along the hot rotational poles, flows toward the equator at the surface, and sinks inward at the equator on the Eddington–Sweet timescale (t_ES ≈ t_KH [ GM / (Ω² R³) ]), dredging CNO-processed nitrogen-14 and helium from the core to the photosphere of massive OB stars and extending their main-sequence lifetimes (Maeder & Meynet, 2000).

1. von Zeipel, H. (1924). "The radiative equilibrium of a slightly oblate rotating star." Monthly Notices of the Royal Astronomical Society, 84(9), 665–683. https://doi.org/10.1093/mnras/84.9.665
2. Lucy, L. B. (1967). "Gravity-darkening for stars with convective envelopes." Zeitschrift für Astrophysik, 65, 89–92. https://ui.adsabs.harvard.edu/abs/1967ZA.....65...89L
3. Monnier, J. D., Zhao, M., Pedretti, E., Thureau, N., Ireland, M., Muirhead, P., et al. (2007). "Imaging the Surface of Altair." Science, 317(5836), 342–345. https://doi.org/10.1126/science.1143205
4. Maeder, A. (2009). Physics, Formation and Evolution of Rotating Stars. Springer. https://doi.org/10.1007/978-3-540-76949-1
5. Espinosa Lara, F., & Rieutord, M. (2011). "Gravity darkening in rotating stars." Astronomy & Astrophysics, 533, A43. https://doi.org/10.1051/0004-6361/201117252