Kerr Perturbations and Teukolsky Theory

Kerr rotation couples metric components, angular structure, and boundary flux in ways that obscure the two radiative degrees of freedom. The Teukolsky framework resolves this by replacing most metric variables with extreme Newman–Penrose curvature, exploiting type-D principal null directions, and separating the resulting equation into radial and angular problems. This chapter develops that chain from tetrad choice to sourced curvature and carefully bounded metric reconstruction. Exact Kerr geometry is input; self-consistent orbital evolution remains outside the chapter.

Helpful background. Black-Hole Geometry and Mechanics supplies the Kerr exterior and horizon generator; PDEs and Green Functions for Gravitation supplies separation, Green functions, and distributions; Formulations of General Relativity supplies gauge and sourced linearized equations.

Curvature variables on a rotating background

The central move is to align a complex null tetrad $(\ell,n,m,\bar m)$ with Kerr’s repeated principal null directions. With the site’s signature $(-,+,+,+)$, this chapter uses $\ell\!\cdot n=-1$ and $m\!\cdot\bar m=1$. In a Kinnersley tetrad the background has only $\Psi_2\ne0$ among its Weyl scalars. Consequently, the first-order extreme components $\delta\Psi_0$ and $\delta\Psi_4$ isolate radiative curvature and obey decoupled equations.

Every page uses

$$\Delta=r^2-2GMr+a^2, \qquad e^{-i\omega t+im\phi},$$

where $M$ is physical mass and $a=J/M$ has dimensions of length. The future horizon generator is $\chi_H=\partial_t+\Omega_H\partial_\phi$, so its characteristic mode frequency is $k_H=\omega-m\Omega_H$. A future-horizon condition must be interpreted in a horizon-penetrating chart and regular tetrad. A power of $\Delta$ in a Kinnersley component is not by itself a curvature singularity.

The Weyl and field definitions important to this chapter are

$$\begin{aligned} \rho_{NP}&=-\frac1{r-ia\cos\theta}, &\Psi_2&=\frac{GM}{(r-ia\cos\theta)^3},\\ \psi_{+2}&=\delta\Psi_0, &\psi_{-2}&=\rho_{NP}^{-4}\delta\Psi_4. \end{aligned}$$

The two master fields have different length dimensions, $L^{-2}$ and $L^2$. Neither equality of raw radial coefficients nor a bare spin-sign change equates their physical content. Where a leaf abbreviates geometric mass by $\mu$ or $M_g$, it means $GM$; $\mu_{\mathrm{NP}}$ is instead a connection coefficient, and $\mu_p$ is the physical particle mass.

Teukolsky’s original paper uses the opposite metric signature and geometrized units. The metric, connection and Weyl definitions are translated together, including the positive background $\Psi_2$ above Teukolsky 1973, § II, pp. 637–638 and Eqs. (4.4)–(4.7), p. 640, PDF. Amplitudes, eigenvalues and sources are compared only after fixing their complete definitions.

The six-page route

The pages are arranged in the order needed to build a controlled calculation:

1. Kinnersley Tetrads and Newman–Penrose Curvature fixes null normalization, all twelve connection coefficients, background Weyl curvature, residual spin-boost freedom, and horizon-regular alternatives.
2. Derivation of the Teukolsky Equation carries out the two-Bianchi/one-Ricci elimination for the +2 sector, including its weighted commutator, and separates the translated coordinate operator.
3. Radial Teukolsky Equation and Physical Solutions derives the endpoint powers, distinguishes the Green-function Wronskian from the energy relation, and gives the full real-frequency gravitational flux conversion. It uses the separation constant defined on the angular page.
4. Angular Teukolsky Equation and Separation Constants defines spin-weighted spheroidal harmonics, pole regularity, eigenvalue shifts, spherical limits, and the complex-frequency adjoint problem.
5. Teukolsky Sources and Point-Particle Coupling derives a complete circular-equatorial source, its delta-derivative coefficients and normalized waveform, then checks its signed weak-field quadrupole limit.
6. Hertz Potentials and Metric Reconstruction applies an explicit IRG operator to a local static quadrupole, recovers its curvature from the resulting metric, and exhibits the missing mass/spin information. It then separates the additional source, gauge and global-boundary questions.

The order is conceptual, not compulsory. A reader who already knows the Newman–Penrose formalism can enter at the decoupling derivation. A flux calculation needs the tetrad, radial, angular, and source pages but may postpone metric reconstruction. A second-order or self-force reader should follow the full route, because the final page explains exactly which information curvature alone omits.

Three reader routes

Equation route. Read pages 1–4 in order. This route explains why separation works, how $s$ enters the master operator, why the angular eigenvalue depends on $a\omega$, and how radial boundary data select a physical solution. It is the shortest preparation for Kerr scattering and quasinormal boundary conditions.

Source-and-flux route. Read pages 1, 3, 4, and 5. Use polar norm $\int|S|^2\sin\theta\,d\theta=1$ and the azimuthal basis $S e^{im\phi}$, whose full-sphere norm squared is $2\pi$. Project the conserved particle stress tensor with the chapter’s $d\omega/(2\pi)$ Fourier measure, solve the positive-source radial equation, and convert its amplitudes to fluxes Sasaki–Tagoshi 2003, § 2.1, Eqs. (13)–(36), pp. 9–12, PDF. The complete source formula, not an isolated printed coefficient, fixes this normalization.

Metric route. Read pages 1–2 and 6, returning to pages 3–4 for mode inversion. This route distinguishes an adjoint Teukolsky solution from a Hertz potential chosen to reproduce a specified Weyl scalar, and a reconstructed radiation-gauge piece from a completed metric. Wald’s adjoint identity supplies the abstract construction Wald 1978, full article, pp. 203–206; Chrzanowski gives the explicit Kerr reconstruction operators Chrzanowski 1975, full article, pp. 2042–2062.

A recurring circular-orbit calculation

A point mass on a circular equatorial geodesic provides one calculation that crosses the chapter. Its stress tensor is conserved on the fixed Kerr background. Projection onto $e^{im\phi}$ and Fourier transformation with $e^{-i\omega t}$ restrict the eternal source to $\omega=m\Omega$. Angular regularity selects a spin-weighted spheroidal harmonic and its shifted radial constant. The prepared source is a finite combination of $\delta(r-r_0)$ and its first two derivatives.

The retarded radial solution uses an in solution at the horizon and an up solution at infinity. Its source integral reduces to homogeneous functions and their first two derivatives at $r_0$, divided by the correctly normalized weighted Wronskian. The source page checks the resulting weak-field waveform against the direct Newtonian quadrupole wave, including the sign that squared power alone cannot test.

The fixed geodesic has constant orbital energy. Its computed radiative fluxes instead supply the leading orbit-averaged dissipative loss when the orbit is promoted to an adiabatically evolving two-body approximation:

$$\left\langle\dot E_{\mathrm{orb}}\right\rangle +\mathcal F_E^\infty+\mathcal F_E^H=0.$$

A negative $\partial_t$ Killing-energy flux into the horizon is possible when $\omega>0$ and $k_H\lt0$, and represents rotational energy extraction. The horizon-generator energy remains nonnegative. The homogeneous scattering balance follows from an opposite-spin conjugated Wronskian, not by interpreting the same-spin Green-function Wronskian as an energy current Teukolsky–Press 1974, § V, Eqs. (5.2)–(5.8), p. 452, PDF.

If a metric is required, reconstruct the vacuum regions inside and outside the orbit in a declared radiation gauge, determine the mass and angular-momentum completion from charges or matching, and handle source-induced strings or jumps explicitly. The classic vacuum formula does not yield a single globally smooth metric through the particle. Modern sourced schemes add a corrector carrying stress-tensor components incompatible with pure radiation gauge Hollands–Toomani 2024, §§1.1–1.3 and §3.

Distinctions that control the answer

Keep the following pairs separate throughout the chapter:

• Coordinate gauge versus tetrad gauge: $h_{ab}\mapsto h_{ab}+\mathcal L_Xg_{ab}$ differs from a spin-boost of $(\ell,n,m,\bar m)$.
• Separation versus boundary conditions: a product ansatz yields ODEs; it does not select retarded, scattering, quasinormal, or advanced solutions.
• Weyl amplitude versus flux: a raw $R_{s\ell m\omega}$ coefficient depends on field normalization; physical flux includes reconstruction and conversion factors.
• Angular $\Lambda$ versus radial $\lambda$: they differ by $-s+(a\omega)^2-2am\omega$ in this chapter’s dictionary.
• Vacuum reconstruction versus sourced metric: classic Hertz operators solve the homogeneous Einstein equation in a radiation gauge; generic sources require regional matching or a corrector.
• Radiative curvature versus complete metric: the explicit mass/spin parameter examples change charges without changing either extreme curvature. They prove missing information, not a kernel classification on every possible domain.
• Test-body source versus inspiral dynamics: a prescribed geodesic sources first-order curvature; radiation reaction and self-consistent evolution require a separate two-body approximation.

This chapter formulates the continuum Kerr perturbation problem, specifying the conventions needed to reproduce calculations and the limits of metric reconstruction. The geometry volumes develop exact Kerr extensions and horizon geometry. Numerical algorithms, waveform construction, and self-consistent inspiral modeling require additional tools beyond this continuum description.

Two further limits keep the examples consistent. The derivation’s fixed-spin anti-linear map transforms both separated operators; the angular page’s opposite-spin identity concerns its angular equation alone. Neither is an automatic equivalence of retarded boundary problems. Likewise, the local static Hertz example is an externally applied tide, not an isolated wave or a zero-frequency use of the scattering flux formulas. Adding a mass or spin completion does not automatically preserve its radiation gauge.

Chapter review

Consider a proposed $s=-2$ calculation that (i) imports Teukolsky’s tetrad products without translating signature, (ii) calls $e^{+ik_Hr_*}$ future-horizon ingoing, (iii) uses the ordinary Wronskian $R_1R_2'-R_2R_1'$ as a constant, (iv) projects an eternal fixed circular particle source but obtains a continuum of real frequencies, and (v) declares the ORG metric globally unique and complete because it reproduces $\delta\Psi_4$. Diagnose each step. Finally, replace $B=S e^{im\phi}$ by the unit-sphere harmonic $Y=B/\sqrt{2\pi}$: how must the field and source coefficients change?

Solution

For (i), Teukolsky uses $+---$, whereas this chapter uses $-+++$. Recompute or translate the metric, null products, Weyl-scalar definitions, spin coefficients, and source contractions together. A scalar $s=0$ reduction and direct tetrad inner products provide independent checks.

For (ii), with $e^{-i\omega t+im\phi}$ the future-horizon ingoing phase is $e^{-ik_Hr_*}$ when expressed with an advanced co-rotating coordinate; spin-dependent powers of $\Delta$ must then be converted to a regular tetrad. The opposite exponential is the outgoing local branch.

For (iii), the radial equation’s derivative term is in weighted divergence form. The constant object is

$$W_s=\Delta^{s+1}(R_1R_2'-R_2R_1'),$$

not the unweighted bracket. Evaluating $W_s$ at both boundaries checks the phases and normalization.

For (iv), an eternal circular equatorial orbit has $\phi_p=\Omega t+\phi_0$. The time integral is $2\pi e^{-im\phi_0}\delta(\omega-m\Omega)$. A continuum signals a Fourier-phase error unless the orbit has finite duration or evolves, which must be stated as a new approximation.

For (v), reproducing one Weyl scalar checks the radiative round trip only in the chosen vacuum mode domain. ORG fixes a gauge representative but need not be regular through a source or on every boundary. Add independently determined $\delta M$ and $\delta J$, allow a residual gauge term, and use regional matching or a sourced corrector. State the source class, domain, boundary data, regularity, and charge conditions before making any uniqueness claim.

For the last question, multiply both radial and source coefficients by $\sqrt{2\pi}$, since $RB=(\sqrt{2\pi}R)Y$. An outgoing discrete amplitude therefore becomes $Z_{\rm unit}=\sqrt{2\pi}Z$, and

$$\frac{|Z_{\rm unit}|^2}{4\pi G\omega^2} =\frac{|Z|^2}{2G\omega^2}.$$

The actual curvature field and its flux are unchanged. Leaving the coefficient fixed would change the field, not merely its notation.

References

• Chrzanowski, Paul L. “Vector Potential and Metric Perturbations of a Rotating Black Hole.” Physical Review D 11 (1975): 2042–2062. DOI.
• Hollands, Stefan, and Vahid Toomani. “Metric Reconstruction in Kerr Spacetime.” Classical and Quantum Gravity 42 (2025): 175008. DOI. Open PDF.
• Sasaki, Misao, and Hideyuki Tagoshi. “Analytic Black Hole Perturbation Approach to Gravitational Radiation.” Living Reviews in Relativity 6 (2003): article 6. DOI. Open PDF.
• Teukolsky, Saul A. “Perturbations of a Rotating Black Hole. I. Fundamental Equations for Gravitational, Electromagnetic, and Neutrino-Field Perturbations.” Astrophysical Journal 185 (1973): 635–647. DOI.
• Teukolsky, Saul A., and William H. Press. “Perturbations of a Rotating Black Hole. III. Interaction of the Hole with Gravitational and Electromagnetic Radiation.” Astrophysical Journal 193 (1974): 443–461. ADS record and PDF.
• Wald, Robert M. “Construction of Solutions of Gravitational, Electromagnetic, or Other Perturbation Equations from Solutions of Decoupled Equations.” Physical Review Letters 41 (1978): 203–206. DOI.