Black-Hole and Gravitational-Wave Probes

Black-hole observations test propagation, binary dynamics, and remnant structure in the strongest accessible gravitational fields. They do not automatically test microscopic quantum gravity. An echo, a shifted quasinormal frequency, or a nonzero tidal response can also arise from an ordinary horizonless compact object, waveform mismatch, or detector noise. The scientific task is to map a controlled microscopic correction to a normalized detector observable and compare it with those alternatives.

Required background. Equivalence Principles and Gravitational Measurement supplies operationally meaningful timing, phase, and tidal observables.

Helpful background. Flat-Spacetime Quantum Field Theory supplies effective operators; Einstein Equations supplies Kerr and compact-binary dynamics; Quantum Information supplies model comparison; Quantum Gravity supplies claim boundaries; and Gravitational Radiation and Waveforms supplies detector response and parameter estimation.

Signals and their inference targets

A general frequency-domain signal model can be organized as

$$\tilde h_I(f) =R_I(f)\, \left[ \tilde h_{\mathrm{GR}}(f;\vartheta) \exp[i\,\delta\Psi(f;\lambda)] +\delta\tilde h_{\mathrm{rem}}(f;\lambda) \right] +\tilde n_I(f),$$

where $R_I$ is detector $I$‘s calibrated response, $\vartheta$ denotes source parameters, and $\lambda$ denotes a declared departure. The three contributions have different meanings: a phase correction to generation or propagation, a change in merger or remnant radiation, and detector noise. A valid analysis specifies which term contains the proposed physics rather than letting one flexible residual absorb all three.

For a Kerr remnant, a parameterized ringdown test writes

$$f_{\ell mn} =f_{\ell mn}^{\mathrm{Kerr}}(M_f,\chi_f) (1+\delta\widehat f_{\ell mn}), \qquad \tau_{\ell mn} =\tau_{\ell mn}^{\mathrm{Kerr}}(M_f,\chi_f) (1+\delta\widehat\tau_{\ell mn}).$$

A measured nonzero deviation would falsify that Kerr template under the chosen mode content and start time. It would not identify quantum gravity: environmental matter, nonlinear or missing modes, an exotic compact object, and waveform error can shift the same fitted quantities.

Two other popular observables illustrate the same degeneracy. If a reflecting surface lies at

$$r_0=2GM(1+\epsilon),$$

the Schwarzschild tortoise distance gives the leading echo-delay scaling

$$\Delta t_{\mathrm{echo}} \simeq4GM\lvert\ln\epsilon\rvert$$

for $0\lt\epsilon\ll1$, up to barrier, spin, and boundary-phase corrections. Here $M$ is the source-frame mass and the time is measured by a distant observer in the source’s asymptotically flat frame. A cosmological detector measures $(1+z)\Delta t_{\mathrm{echo}}$, equivalently using the detector-frame mass $(1+z)M$. This delay fixes a redshifted location only after the reflectivity and transfer function are supplied. For tides,

$$Q_{ij}=-\lambda_T\mathcal E_{ij}, \qquad \lambda_T=\frac{2}{3}\frac{k_2R^5}{G}.$$

Here the tidal formula is the static quadrupolar response of a nonrotating spherical object, with $R$ its areal radius and $\mathcal E_{ij}$ the applied electric tidal tensor. A nonzero conservative $k_2$ distinguishes that object from a four-dimensional Schwarzschild black hole in vacuum GR under the same response convention; ordinary material stars and many horizonless models also have $k_2\ne0$. Rotating and time-dependent response requires spin, frequency, and response components to be specified rather than using this one scalar formula unchanged. Echoes, Love numbers, multipolar structure, and images therefore diagnose compact-object structure, not its microscopic origin by themselves Cardoso and Pani 2019, §§4–7.

The pinned GWTC-5.0 data contract

This discussion reflects results available through 2026-08-20. The current official product is GWTC-5.0, including observations through O4b. The source strain is the GWOSC event release, DOI 10.7935/bk00-6a89, with 4096-second event files at 16384 Hz and a separately downsampled 4096 Hz series, event data-quality and injection masks, and Stable Release Version 9 parameter-estimation samples. The version-9 candidate record is Zenodo 10.5281/zenodo.20348004 and the two parameter-estimation parts are 10.5281/zenodo.20348005 and 10.5281/zenodo.20348006 GWOSC 2026, GWTC-5.0 release documentation.

The dedicated tests-of-GR sample is narrower than the catalog’s probable-event list. It requires a false-alarm rate no larger than $10^{-3}\,\mathrm{yr}^{-1}$ and observation by at least two detectors. It contains 72 O4b events, five newly qualifying reanalysed O4a events, and earlier events, for 168 total. Each test imposes additional signal-strength, mass, frequency, or mode selections; no result should silently substitute the catalog-wide $p_{\mathrm{astro}}\gt0.5$ sample. These selections and the event-by-test matrix are given in LIGO–Virgo–KAGRA 2026, §§1–2 and tables 2–3.

Calibration. Use the event-specific GWOSC strain and data-quality version. Frequency-dependent response amplitude and phase are spline nuisance functions with Gaussian calibration priors and are marginalized with source parameters. GW240925_005809 specifically uses the C01 recalibration; treating its earlier response as interchangeable would bias a high-precision test LIGO–Virgo–KAGRA 2026, §§1 and 3.1.

Likelihood and priors. For stationary Gaussian noise with one-sided power spectrum $S_{n,I}$,

$$\ln\mathcal L(d\mid\theta,M) =-\frac12\sum_I \langle d_I-h_I(\theta)\mid d_I-h_I(\theta)\rangle_I +\text{constant},$$

with

$$\langle a\mid b\rangle_I =4\operatorname{Re}\int_{f_{\min}}^{f_{\max}} \frac{\tilde a(f)\tilde b^*(f)}{S_{n,I}(f)}\,df.$$

The catalog analysis uses Bilby/BilbyTGR, usually Dynesty, weakly informative compact-binary priors, test-specific deviation priors, and hierarchical Gaussian population models for deviations that may vary across events. A reproduction must use each test’s published prior ranges and event selection, not merely the public GR posterior samples.

What GWTC-5.0 does and does not find

The official paper reports seven complementary analyses. The data products needed to regenerate its figures and tables are released as LIGO-P2600410-v1, with separate archives for residuals, polarization, FTI/TIGER generation tests, pSEOB ringdown, time-domain ringdown, and QNM rational filtering LVK 2026, tests-of-GR data release.

Channel	Pinned waveform or statistic	Cutoff result and interpretation
Coherent residuals	Best-fit GR subtraction followed by BayesWave noise-background comparison	Residuals are consistent with detector noise; a generic residual test is less sensitive than a targeted deformation
Extra polarizations	Bayes factors for tensor, vector, and scalar response hypotheses	No strong evidence for polarization content beyond GR
Inspiral generation	TIGER with IMRPhenomXPHM SpinTaylor; FTI with SEOBNRv5HM ROM	Parameterized post-Newtonian bounds improve by factors 1.2–2.6 and remain consistent with GR
Remnant spectrum	Time-domain ringdown, pSEOBNRv5PHM full-signal test, and QNMRF	Overall Kerr consistency; catalog-combination tails and event-specific high-frequency content require caution

The paper’s conclusion is no significant evidence for physics beyond GR within these tests LIGO–Virgo–KAGRA 2026, §8. That is not a model-independent exclusion of quantum gravity. The parameterized coefficients are altered by many classical theories, and a microscopic model generally changes several correlated coefficients rather than one at a time.

The omissions are equally informative. The GWTC-5.0 tests paper discontinues catalog echo searches because they were not expected to improve earlier constraints; it does not thereby prove that every surface model is absent. A dispersive-propagation test was withheld while correlations and event selection were being refined. Tidal response requires long inspirals and an appropriate material or compact-object model. Population combination requires a selection function and hyperprior; multiplying posteriors from a handpicked set of anomalous events is invalid.

Systematics and discriminating checks

Every claim must vary the noise spectrum, glitch treatment, frequency range, calibration spline, waveform family, higher modes, eccentricity, precession, lensing, ringdown start time, and remnant-mode content. Echo searches additionally require off-source trials over delay and phase; tidal studies require spin and equation-of-state alternatives; population analyses require injection-derived detection efficiency. The loudest event is not automatically the safest if waveform error grows faster than statistical precision.

A genuinely microscopic claim needs more than a nonzero phenomenological parameter. The same coupling should predict a correlated pattern across inspiral, merger, ringdown, or an independent channel, remain within its effective-field-theory regime, and fail a named conventional alternative. A generic exotic compact object is the null competitor for horizon-scale effects. A wrong mass scaling, inconsistent mode relation, or disappearance under a second waveform family is adverse evidence.

Reproduction status. No GWTC-5.0 strain likelihood, posterior, or injection campaign was rerun for this page. The official statistics and products are quoted with their versions. A clean rerun would pin the event files and calibration, verify data-quality segments and hardware injections, reproduce one released figure from LIGO-P2600410-v1, then repeat the inference with a null injection and an alternate waveform. Anything less is a source comparison, not an independent reproduction.

Exercise

An echo model places a surface at $r_0=2GM(1+\epsilon)$ and predicts $\Delta t_{\mathrm{echo}}/(4GM)=80$. Estimate $\epsilon$. Does measuring this delay identify a quantum surface?

Solution

The leading relation gives

$$\lvert\ln\epsilon\rvert=80, \qquad \epsilon=e^{-80}\approx1.8\times10^{-35}.$$

This infers an extremely redshifted effective location within the assumed Schwarzschild, reflecting-boundary model. The same delay can arise from different reflectivity phases, potential barriers, spin corrections, or horizonless compact objects. A quantum-gravity interpretation would require a microscopic calculation of the boundary response and correlated predictions beyond the delay, plus trials-calibrated evidence against glitches and waveform residuals.

References

• Cardoso, Vitor, and Paolo Pani (2019), “Testing the Nature of Dark Compact Objects: A Status Report,” Living Reviews in Relativity 22, 4, doi:10.1007/s41114-019-0020-4.
• Gravitational Wave Open Science Center (2026), “GWTC-5.0 Data Release Documentation,” version released 26 May 2026, doi:10.7935/bk00-6a89.
• LIGO Scientific Collaboration, Virgo Collaboration, and KAGRA Collaboration (2026), “Data Release for GWTC-5.0: Tests of General Relativity,” LIGO Document P2600410-v1, official record.
• LIGO Scientific Collaboration, Virgo Collaboration, and KAGRA Collaboration (2026), “GWTC-5.0: Tests of General Relativity,” preprint, arXiv:2607.19293.