Effective-One-Body Waveforms

An effective-one-body (EOB) waveform is an inspiral–merger–ringdown construction built around a mapped one-body Hamiltonian. Its strength is architectural: weak-field information, strong-field resummation, dissipative driving, multipolar radiation, merger attachment, and numerical calibration remain identifiable rather than being collapsed into one fit. This page explains that assembly. Relativistic Two-Body Dynamics supplies the conservative and dissipative inputs; production software and numerical-relativity algorithms remain with Computational Relativity; detector-specific inference verdicts remain downstream.

Helpful background. Einstein Equations supplies the field equations behind the weak-field inputs; Waves, Green Functions, and Geometric Optics supplies radiation and Fourier methods; Relativistic Two-Body Dynamics supplies Hamiltonian and flux information; Computational Relativity supplies numerical-relativity calibration waveforms; and Gravitational-Wave Astronomy treats detector-specific likelihoods and inference claims.

The real–effective energy map

Let $M=m_1+m_2$, $\mu=m_1m_2/M$, and $\nu=\mu/M$. A central nonspinning EOB dictionary relates the real Hamiltonian to a dimensionless effective Hamiltonian $h_{\mathrm{eff}}=H_{\mathrm{eff}}/\mu$ by

$$H_{\mathrm{EOB}} =M\sqrt{1+2\nu\left(h_{\mathrm{eff}}-1\right)}.$$

Buonanno and Damour obtain this square-root map by matching the real and effective energy spectra through the available post-Newtonian order Buonanno and Damour 1999, §IV, especially Eqs. (4.25)–(4.26), PDF. Once adopted, the displayed relation is an exact algebraic dictionary inside the model. It is not an exact all-orders theorem about the two-body problem: $H_{\mathrm{eff}}$, its potentials, and the matching information remain finite-order inputs.

Writing $\delta=h_{\mathrm{eff}}-1$ gives the small-$\nu$ expansion

$$H_{\mathrm{EOB}}-M =\mu\delta- \frac{\mu\nu}{2}\delta^2+O(\nu^2\mu).$$

Thus the test-mass limit recovers the effective particle’s binding energy, while finite-$\nu$ terms encode the energy map between the two descriptions. The weak-field limit additionally requires $\delta\to0$ as $GM/r\to0$.

Conservative dynamics and resummation

The effective Hamiltonian describes motion in a $\nu$-deformed black-hole geometry, supplemented at higher order by nongeodesic momentum terms. At any fixed post-Newtonian order, the radial potentials are finite series in $u=GM/r$. EOB models reorganize those series—through rational forms, logarithmic resummations, or factored expressions—to preserve useful strong-field structure and avoid obvious pathologies.

That reorganization is a model choice. Two potentials with the same expansion through $u^N$ can differ at $u^{N+1}$ and beyond, just as two Taylor approximants can. Their circular-orbit energy and frequency must reproduce the supplied PN expansion, and their $\nu\to0$ limit must recover the appropriate test-particle dynamics. Buonanno and Damour develop the effective metric and real Hamiltonian through this sequence in Buonanno and Damour 1999, §§II–VI, PDF.

Modern aligned-spin implementations add spin-orbit and spin-spin terms, resum selected potentials, and introduce a small number of calibration functions. Those functions are not interchangeable with known analytic coefficients: every calibrated term should be named separately from terms fixed by PN, self-force, or perturbation theory.

Radiation reaction and inspiral–plunge modes

Hamilton’s conservative equations are driven by a radiation-reaction force constructed from an outward waveform flux. Schematically,

$$\dot p_\phi=\mathcal F_\phi, \qquad \mathcal F_\phi\simeq-\frac{\mathcal F_E}{\Omega},$$

for a quasi-circular orbit, with the minus sign expressing source angular-momentum loss. The balance relation has a controlled adiabatic interpretation during inspiral; close to merger, the force and waveform are model ingredients checked against perturbative and numerical results. Buonanno and Damour’s original transition from conservative motion to radiation-driven plunge is described in Buonanno and Damour 1999, §VII, PDF.

With $h=h_+-ih_\times$, unit-normalized spin-weight $-2$ harmonics, and the inherited $e^{-i\omega t}$ convention,

$$h(u,\theta,\phi) =\sum_{\ell,m}h_{\ell m}(u)\,{}_{-2}Y_{\ell m}(\theta,\phi).$$

Inspiral–plunge modes are commonly factorized into a Newtonian leading factor, an effective source, resummed tail and phase factors, residual amplitudes, and a non-quasicircular (NQC) correction. NQC factors repair amplitude and frequency behavior when radial motion ceases to be negligible; they are not a derivation of merger from PN theory. Pompili and collaborators give the factorized modes, added multipoles, NQC conditions, and radiation reaction for the dated SEOBNRv5HM construction in Pompili et al. 2023, §§II–III and Appendices A–C, PDF.

The feedback is important: the multipolar waveform estimates the flux, the flux drives the orbit, and that orbit supplies the next waveform phase and amplitude. A flux used in the dynamics may retain a different mode set or approximation from the final output waveform, but that choice must be stated and tested. Horizon absorption, when included, also contributes to the orbital energy budget even though it is not outgoing radiation at infinity. This is why a waveform model needs more information than a conservative Hamiltonian and a late-time fit.

Merger–ringdown attachment

A useful simplified attachment model treats each spherical-harmonic mode as one spheroidal family and neglects spherical–spheroidal mixing. In that approximation an inspiral–plunge mode is attached near a declared matching time to a finite damped sum,

$$h_{\ell m}^{\mathrm{RD}}(u) =\sum_{n=0}^{N}A_{\ell mn} e^{-i\omega_{\ell mn}(u-u_{\mathrm{match}})}, \qquad \operatorname{Im}\omega_{\ell mn}\lt 0.$$

The complex amplitudes are fixed by specified matching conditions or fitted templates; the frequencies depend on the remnant mass and spin. This formula is not a basis identity: a spherical $h_{\ell m}$ generally receives projections from several spheroidal $\ell'$ families at fixed $m$. Attachment time, overtone content, mixing prescription, mirror branches, and continuity order are therefore part of the model definition. Continuity of $h$ alone does not guarantee continuity of $\dot h$, and an apparently smooth attachment does not establish physical accuracy.

SEOBNRv5HM does not use the displayed finite-overtone sum as its general merger–ringdown ansatz. For its monotonic modes it factors a phenomenological amplitude-and-phase form by the least-damped quasinormal mode; it treats the $(3,2)$ and $(4,3)$ spherical modes with explicit spherical–spheroidal mixing Pompili et al. 2023, §III, Eqs. (48)–(56), PDF. Its reported performance is conditional on that paper’s quasi-circular, nonprecessing binary-black-hole domain, listed modes, numerical-relativity set, alignment prescription, frequency band, and noise curves. It should not be generalized to eccentric, precessing, or uncalibrated regions without a separate test.

What calibration can and cannot establish

Calibration adjusts designated coefficients against numerical-relativity waveforms while holding the analytic architecture fixed. Validation then uses simulations or parameter-space points not used for fitting, including boundaries and sparse corners. Pompili et al. state the 442-waveform NR calibration set, additional perturbative information, fit construction, and model comparisons in Pompili et al. 2023, §§IV–V and Appendices D–F, PDF. Those numbers describe the 2023 SEOBNRv5HM release, not an immutable EOB domain.

A useful error decomposition distinguishes analytic truncation and resummation, calibration fit error, NR truncation and extraction error, attachment error, and omitted-mode or omitted-physics error. A small mismatch in one validation set cannot identify which component dominates. If matches are quoted, the one-sided PSD, frequency band, included modes, source or inertial frame, alignment, and time/phase maximization must accompany the number.

Common pitfalls

Calling EOB purely analytic. Its organizing map and many coefficients are analytic, but contemporary waveform families usually contain calibrated potentials, NQC fits, remnant fits, and attachment choices.

Equating smoothness with correctness. A $C^1$ attachment can still use the wrong remnant parameters or omit important mixed modes. Smoothness is a necessary implementation check, not a physics validation.

Extending a version label beyond its domain. SEOBNRv5HM denotes a specific dated construction. Accuracy statements inherit its parameter range, modes, waveform convention, numerical references, and comparison norm.

Exercise

Let $\delta=h_{\mathrm{eff}}-1=-0.02$ and $\nu=1/4$. Expand the EOB energy map through the first finite-$\nu$ correction and compute $(H_{\mathrm{EOB}}-M)/\mu$. Compare with the unexpanded square root.

Solution

The expansion gives

$$\frac{H_{\mathrm{EOB}}-M}{\mu} =\delta-\frac{\nu}{2}\delta^2 =-0.02-\frac18(0.0004) =-0.02005.$$

The unexpanded result is

$$\frac{\sqrt{1+2\nu\delta}-1}{\nu} =4\left(\sqrt{0.99}-1\right) \simeq-0.02005025.$$

The difference, about $2.5\times10^{-7}$, begins at the next order in the expansion. The calculation checks the test-mass interpretation and also shows why re-expanding versus retaining the square root is a controlled model choice at finite order.

References

• Buonanno, Alessandra, and Thibault Damour. “Effective One-Body Approach to General Relativistic Two-Body Dynamics.” Physical Review D 59 (1999): 084006. DOI. Open PDF.
• Pompili, Lorenzo, et al. “Laying the Foundation of the Effective-One-Body Waveform Models SEOBNRv5: Improved Accuracy and Efficiency for Spinning Nonprecessing Binary Black Holes.” Physical Review D 108 (2023): 124035. DOI. Open PDF.