Kompaneets Equation

The Kompaneets Equation is the fundamental partial differential kinetic equation of radiative transfer and high-energy astrophysics describing how the phase-space photon occupation number n(ν, t) of an isotropic radiation field evolves through repeated, small-energy-transfer Compton scattering (Comptonization) off a thermal Maxwell–Boltzmann distribution of non-relativistic electrons. Derived by the Soviet theoretical physicist Aleksandr Solomonovich Kompaneets in 1949 inside the Soviet nuclear program and published in the open scientific literature in 1956 (following earlier unpublished wartime work by Ray Weymann and J. Robert Oppenheimer's group), the equation takes the form of a nonlinear Fokker–Planck diffusion equation in dimensionless photon frequency space x = hν / (k_BT_e). Because pure Compton scattering conserves total photon number while exchanging thermal energy between electrons and radiation, the Kompaneets equation underpins the physics of X-ray binary and active galactic nucleus (AGN) coronae, early-universe Bose–Einstein μ-distortions, and the thermal Sunyaev–Zeldovich effect in the cosmic microwave background (CMB).

Consider an isotropic, spatially homogeneous gas of free electrons with number density n_e and non-relativistic thermal temperature T_e (k_BT_e ≪ m_ec² = 511 keV) interacting with an isotropic radiation field of low-energy photons (hν ≪ m_ec²). Let n(ν, t) denote the dimensionless Bose photon occupation number (mean number of photons per quantum phase-space cell h³ of a single polarization state), related to the specific radiative intensity I_ν and spectral energy density u_ν by:

I_ν = (2*hν*³ / c²) n(ν, t), u_ν = (8*πhν*³ / c³) n(ν, t)

In a single elastic scattering event γ + e⁻ → γ′ + e⁻′, a photon of frequency ν scatters off an electron moving with velocity v (v / c ~ √(k_BT_e / m_ec²) ≪ 1). Two competing mechanisms shift the photon's frequency by a small fractional amount Δν / ν ≪ 1:

1. Quantum Electron Recoil (Downward Drift): Even if the electron is initially at rest, the photon transfers momentum Δp ~ hν / c to the electron, losing kinetic recoil energy Δν_recoil / ν = −(hν / m_ec²)(1 − cos θ), where θ is the scattering angle.
2. Thermal Doppler Shifts (Random Walk + Upward Drift): In the laboratory frame, scattering off an electron moving with velocity v produces a first-order linear Doppler shift Δν_D,1 / ν ~ ±(v / c) with zero isotropic mean but nonzero second-moment variance ⟨(Δν / ν)²⟩ = (2k_BT_e / m_ec²)(1 − cos θ), plus a second-order relativistic aberration boost ⟨Δν / ν⟩_D,2 = +4(k_BT_e / m_ec²).

Fokker–Planck Expansion and Photon Number Conservation

Because the fractional frequency shift per Thomson collision is small (|Δν / ν| ≪ 1), Kompaneets expanded the full relativistic Boltzmann collision integral in a second-order Taylor series (Fokker–Planck approximation) in dimensionless electron-temperature frequency x ≡ hν / (k_BT_e), explicitly retaining the quantum stimulated-scattering Bose enhancement factors 1 + n(x). Because Compton scattering neither creates nor destroys photons (total photon number density N_γ ∝ ∫₀^∞ x²n(x) dx is conserved), the kinetic equation must take the divergence form of a continuity equation in spherical momentum space x²dx:

∂n / ∂t = (n_e σ_T c) (k_BT_e / m_ec²) (1 / x²) (∂ / ∂x) [ x⁴ ( ∂n / ∂x + n + n² ) ]

Here σ_T = (8π / 3)r_e² = 6.65246 × 10⁻²⁵ cm² is the Thomson scattering cross section. Defining the Compton y-parameter (the dimensionless time coordinate measuring the integrated thermal energy exchange along a photon path dl = c dt):

y ≡ ∫ (k_BT_e / m_ec²) n_e σ_T c dt

simplifies the Kompaneets Equation to its canonical dimensionless form:

∂n / ∂y = (1 / x²) (∂ / ∂x) [ x⁴ ( ∂n / ∂x + n + n² ) ]

Each term inside the radial phase-space flux J_x = −x⁴(∂n / ∂x + n + n²) has a direct physical interpretation:

• ∂n / ∂x represents Doppler frequency diffusion (broadening and upscattering of photons by the thermal random motions of electrons, proportional to T_e).
• n represents spontaneous Compton recoil (downward frequency drift as photons lose energy to recoiling electrons).
• n² represents stimulated Compton recoil (quantum Bose–Einstein enhancement of downward scattering into already populated low-frequency photon states).

Setting the phase-space photon flux to zero (∂n / ∂x + n + n² = 0) yields the most general stationary equilibrium solution of the Kompaneets equation as y → ∞:

n_eq(x) = 1 / [ exp(x + μ) − 1 ] = 1 / [ exp((hν / k_BT_e) + μ) − 1 ]

This is a Bose–Einstein distribution with dimensionless chemical potential μ ≥ 0 at temperature T_r = T_e. Because pure Compton scattering conserves photon number N_γ, a radiation field whose photon density differs from that of a blackbody radiation Planck spectrum at T_e cannot relax to a Planck spectrum (μ = 0) through Compton scattering alone; instead, it settles into a Bose–Einstein spectrum with μ > 0 (or forms a Bose–Einstein condensate spike at x → 0 if there is a photon excess), unless photon-creating processes such as thermal bremsstrahlung (free–free emission) and double Compton scattering (γ + e⁻ ↔ γ + γ + e⁻) drive μ → 0.

Energy Exchange Rate and the Inverse-Compton Temperature

Multiplying the Kompaneets equation by x³ and integrating over all frequencies x ∈ (0, ∞) yields the rate of change of total radiation energy density U_rad ∝ ∫ x³n(x) dx:

dU_rad / dt = 4 n_e σ_T c (k_B / m_ec²) (T_e − T_IC) U_rad

where T_IC is the Compton (or Inverse-Compton) equilibrium temperature of an arbitrary radiation spectrum:

k_BT_IC = (1 / 4) [ ∫₀^∞ hν u_ν (1 + n) dν ] / [ ∫₀^∞ u_ν dν ]

When T_e > T_IC (hot coronal electrons scattering cool seed photons), net energy flows from electrons to photons (inverse Comptonization, cooling the plasma at the Compton cooling rate Λ_C = 4n_eσ_Tc(k_BT_e / m_ec²)U_rad). When T_e < T_IC (hard X-ray/gamma-ray photons illuminating cooler gas), net recoil heats the electrons (Compton heating). For a Planckian blackbody spectrum at temperature T_r, integration by parts shows T_IC = T_r exactly.

In 1969 and 1970, Yakov Zel'dovich and Rashid Sunyaev applied the Kompaneets equation to the cosmic microwave background across cosmic history, discovering the two canonical classes of CMB spectral distortions:

Thermal Sunyaev–Zel'dovich (y-Distortion) Formula

When CMB photons of temperature T_CMB = 2.7255 K pass through hot electrons of temperature T_e ≫ T_CMB (such as the 10⁷–10⁸ K intracluster medium of a galaxy cluster or post-reionization intergalactic gas at z ≲ 1.5 × 10⁴ where y ≪ 1), the dimensionless frequency in electron units is tiny (x_e = hν / k_BT_e ≪ x = hν / k_BT_CMB). Consequently, the Doppler diffusion term ∂n / ∂x_e dominates over recoil n(1 + n) by the factor T_e / T_CMB ≫ 1. Replacing x_e by x = hν / (k_BT_CMB) in the Kompaneets equation gives:

Δn(x) / n₀(x) = y [ x e^x / (e^x − 1) ] [ x coth(x / 2) − 4 ]

This is the classic non-relativistic thermal Sunyaev–Zeldovich effect profile, featuring a decrement in CMB brightness temperature below the null frequency x₀ = 3.830 (ν₀ = 217.6 GHz) and an increment at higher frequencies, increasing total energy density by ΔU_CMB / U_CMB = e^4y − 1 ≈ 4y.

Early-Universe μ-Distortions vs. y-Distortions

In the early universe at redshifts z > z_dC ≈ 2 × 10⁶ (~2 months after the Big Bang), double Compton scattering and thermal bremsstrahlung paired with Compton scattering (y ≫ 1) rapidly restore any injected energy into a pure Planck blackbody at a slightly elevated temperature. Between z ≈ 2 × 10⁶ and z_μ,y ≈ 5 × 10⁴ (where photon creation freezes out but the Compton parameter y > 1 remains large), any energy injection ΔE / E_γ from acoustic wave silk damping, primordial magnetic fields, or decaying particles is driven by the Kompaneets equation into a Bose–Einstein μ-distortion with μ ≈ 1.401 (ΔE / E_γ). Below z ≈ 5 × 10⁴, where y < 1, Comptonization becomes inefficient and energy injections freeze in as a y-distortion (y = (1 / 4) ΔE / E_γ). NASA's COBE/FIRAS instrument constrained both parameters across the full sky to |μ| < 9 × 10⁻⁵ and |y| < 1.5 × 10⁻⁵ (95% CL).

Why does the Kompaneets equation produce a power-law X-ray spectrum in black hole coronae?

In the hot corona above a Shakura–Sunyaev accretion disk (k_BT_e ≈ 50–100 keV, Thomson optical depth τ_T ≈ 0.5–2), soft UV disk photons (hν₀ ~ 10 eV–1 keV) undergo repeated unsaturated inverse-Compton upscattering (y = (4k_BT_e / m_ec²) max(τ_T, τ_T²) ~ 1) while escaping the finite slab at a steady probability per scatter. Balancing the Kompaneets diffusion operator against spatial photon escape −n / t_esc yields a scale-free power-law solution n(x) ∝ x^−(α+3) (specific flux F_ν ∝ ν^−α) with spectral index α = −3/2 + √(9/4 + 4/y), rolling over exponentially above hν ≈ 2–3 k_BT_e where electron recoil halts upscattering.

What is the difference between a Wien spectrum and a Planck blackbody spectrum in the Kompaneets equation?

When photon occupation numbers are very small (n ≪ 1) so stimulated scattering (n²) is negligible, the stationary solution of ∂n / ∂x + n = 0 is the Wien distribution n(x) = C exp(−x), or I_ν ∝ ν³ exp(−hν / k_BT_e), whose mean photon energy is ⟨*hν*⟩ = 3k_BT_e. When n is large enough that stimulated scattering n² is retained and photon-creating processes drive the chemical potential μ → 0, the solution is the Planck blackbody spectrum n(x) = 1 / (exp(x) − 1), whose mean photon energy is ⟨*hν*⟩ ≈ 2.701 k_BT_e.

When does the standard Kompaneets equation break down?

The standard second-order Fokker–Planck approximation breaks down when electrons are relativistic (k_BT_e ≳ 0.1 m_ec² ≈ 50 keV, as in very hot galaxy clusters with k_BT_e ≈ 10–15 keV or pair plasmas) or when photon energies approach the Klein–Nishina regime (hν ≳ m_ec²), because a single scattering event then shifts photon frequency by a large fraction Δν / ν ~ 1. In those regimes, higher-order relativistic derivatives (the Challinor–Lasenby / Itoh relativistic SZ expansions) or the full integral Boltzmann scattering kernel must be used.

1. Kompaneets, A. S. (1956). The establishment of thermal equilibrium between quanta and electrons. Zhurnal Eksperimental'noi i Teoreticheskoi Fiziki, 31, 876–885 [English translation: Soviet Physics JETP, 4(5), 730–737 (1957)]. https://ui.adsabs.harvard.edu/abs/1957JETP....4..730K
2. Weymann, R. (1965). Diffusion approximation for a photon gas interacting with a plasma via the Compton effect. The Physics of Fluids, 8(11), 2112–2114. https://doi.org/10.1063/1.1761165
3. Zel'dovich, Y. B., & Sunyaev, R. A. (1969). The interaction of matter and radiation in a hot-model universe. Astrophysics and Space Science, 4(3), 301–316. https://doi.org/10.1007/BF00661821
4. Sunyaev, R. A., & Zel'dovich, Y. B. (1980). Microwave background radiation as a probe of the contemporary structure and history of the universe. Annual Review of Astronomy and Astrophysics, 18, 537–560. https://doi.org/10.1146/annurev.aa.18.090180.002541
5. Chluba, J., & Sunyaev, R. A. (2012). The evolution of CMB spectral distortions in the early Universe. Monthly Notices of the Royal Astronomical Society, 419(2), 1294–1314. https://doi.org/10.1111/j.1365-2966.2011.19786.x