Timing and Precession Spin Probes

A quasi-periodic oscillation is a measured concentration of variability power, not a labeled geodesic. Spin inference begins only after a physical model identifies the feature with an orbital, radial-epicyclic, vertical-epicyclic, precession, diskoseismic, or instability frequency. A single centroid is usually underdetermined because mass, radius, spin, flow extent, and pressure can trade against one another. The strongest timing case jointly predicts several frequencies, their state evolution, energy-dependent amplitude and phase, and an independently checked mass or radius.

Required background. Probability and Inference for Gravity supplies stochastic-process likelihoods, trials corrections, hierarchical models, and posterior predictive tests.

Helpful background. Black-Hole Geometry and Mechanics supplies Kerr orbital and epicyclic frequencies; Relativistic Matter and Compact Objects supplies oscillating-flow dynamics; and Black-Hole Astrophysics supplies accretion states and spectral-timing context.

The Kerr frequency dictionary

Use radius $r$ in units of $GM/c^2$ and define

$$\nu_0=\frac{c^3}{2\pi GM}.$$

For prograde equatorial circular geodesics with $0\leq a_*\lt1$, the orbital frequency measured at infinity and the two small-oscillation frequencies are given below. For stable or marginally stable radial motion require $r\geq r_{\rm ISCO}(a_*)$; inside the ISCO a negative $\nu_r^2$ signals instability, not an oscillatory frequency.

$$\nu_\phi=\frac{\nu_0}{r^{3/2}+a_*},$$ $$\nu_r^2=\nu_\phi^2 \left(1-\frac{6}{r}+\frac{8a_*}{r^{3/2}}-\frac{3a_*^2}{r^2}\right), \qquad \nu_\theta^2=\nu_\phi^2 \left(1-\frac{4a_*}{r^{3/2}}+\frac{3a_*^2}{r^2}\right).$$

The periastron and nodal precession frequencies are

$$\nu_{\rm per}=\nu_\phi-\nu_r, \qquad \nu_{\rm nod}=\nu_\phi-\nu_\theta.$$

These formulas supply a sharp three-frequency hypothesis: if three observed peaks are respectively $\nu_{\rm nod}$, $\nu_{\rm per}$, and $\nu_\phi$ at one radius, they can determine $(M,a_*,r)$. They do not prove their own identification. Pressure, thickness, magnetic stress, eccentricity, and a radially extended emitting region shift or broaden the fluid modes. Ingram and Motta derive the test-particle dictionary and its disk extension in Ingram and Motta 2019, §§ 4.1–4.4.

For a rigidly precessing inner flow extending from $r_{\rm in}$ to $r_{\rm out}$, the observable is closer to an angular-momentum-weighted average,

$$\nu_{\rm prec} =\frac{\int_{r_{\rm in}}^{r_{\rm out}} \nu_{\rm nod}(r)\,L(r)\,2\pi r\,dr} {\int_{r_{\rm in}}^{r_{\rm out}}L(r)\,2\pi r\,dr},$$

Here $L(r)$ is angular momentum per unit disk area, approximately $\Sigma(r)r^2\Omega(r)$ in the corresponding disk description; the factor $2\pi r\,dr$ supplies the annulus area. This torque-over-angular-momentum approximation assumes a nearly common tilt and sufficiently rapid communication of the warp for the flow to precess coherently. It is not the frequency of every extended tilted flow. A test-particle inversion of this extended-flow frequency can be precise but biased. State evolution can help: a changing $r_{\rm out}$ should move the frequency and also change the spectrum, reverberation scale, and phase-resolved illumination in a linked way.

From events to a QPO likelihood

Start with a named event list and a fixed good-time, detector, energy, and source-state selection. Divide the light curve into equal segments, compute cross-spectra or periodograms with a declared Fourier sign and normalization, and fit the continuum noise and peaks together. For an unaveraged periodogram power $I_j$ with model spectral density $S_j(\Theta)$, the Whittle approximation gives

$$\ln \mathcal L(\Theta) =-\sum_j\left[\ln S_j(\Theta)+\frac{I_j}{S_j(\Theta)}\right],$$

after the sampling window, dead time, Poisson level, and leakage are addressed. Averaged or rebinned powers follow a gamma rather than Gaussian distribution unless enough independent estimates justify the approximation. A Lorentzian peak should report centroid $\nu_c$, width, quality factor $Q=\nu_c/{\rm FWHM}$, fractional rms, coherence, harmonics, energy dependence, and phase lag.

Detection significance must include the searched frequency range, number of observations, states, energy bands, widths, and model alternatives. Choosing the interval because a peak is visible and then quoting a single-bin probability is a look-elsewhere failure. The relevant null is colored accretion noise passed through the same window and search pipeline, not white noise.

The public RXTE/PCA GRO J1655-40 archive analyzed by Motta et al. is a concrete release: 571 observations from 1996–2005, with the observation lists and selected QPO samples in Tables 1–2. The original analysis used PCU2, explicit energy and gain-epoch mappings, HEASoft 6.12, XSPEC 12.7.1, and a $0.6\%$ spectral calibration term; see Motta et al. 2014, §§ 2–3 and Tables 1–2. A reproduction should preserve that configuration and separately compare HEASoft 6.36 plus the currently resolved RXTE CALDB, because channel boundaries, response, dead-time, and background choices are shared calibration systematics.

Results and limitations through 2026-08-18

The evidence does not establish a unique QPO mechanism across all black-hole binaries. The unresolved question is which mode family applies in each source, state, frequency range, and flow geometry.

Evidence or result	Discriminating prediction	Scope or limitation
Kerr geodesics predict a linked orbital, periastron, and nodal frequency triplet	All three features arise at one radius and share one $M$ and $a_*$	A frequency label is assumed; fluid pressure and radial extent are omitted
One RXTE GRO J1655-40 triplet is consistent with the relativistic precession identification and an external dynamical mass	Jointly solve three frequencies, then predict the maximum type-C frequency	It is the only such complete triplet in the RXTE archive and rare-event trials matter
Type-C QPO amplitude, inclination dependence, waveform, and phase-resolved spectra support a geometric modulation	Predict energy-dependent phase, reflection, and harmonic content	Geometry evidence does not uniquely select Lense–Thirring forcing
A precessing torus can explain the same GRO J1655-40 triplet with breathing and vertical modes	Compare the full mode set and its state dependence	More flow parameters admit different mass and spin values
Diskoseismic, accretion–ejection, shock, and empirical-correlation models explain subsets of the phenomenology	Require held-out frequency, phase, and multiwavelength predictions	No alternative has yet passed every source and state; null or missing HFQPOs are common

The triplet calculation and its independent mass comparison are summarized in Ingram and Motta 2019, § 6.4. The same review’s §§ 5.1–5.6 shows why credible competing positions remain: test-particle relativistic precession, a precessing extended flow, corrugation modes, accretion–ejection instability, oscillatory shocks, and pressure or accretion-rate modes. The observed inclination trend and phase-resolved modulation favor geometric motion, but evidence for precession is not automatically evidence for a specific relativistic frequency identification.

The interpretation could change after another stable multi-frequency triplet is found with a predeclared trials analysis; after an independent mass, radius, or inclination measurement conflicts with the timing solution; or after a mode model fails a held-out phase-resolved, state-evolution, or population-scaling prediction.

Adversarial checks and reproduction

Change the mode assignment. Enumerate plausible permutations of the observed peaks and alternative harmonics. Infer $(M,a_*,r)$ under each, rather than conditioning silently on the assignment that gives a familiar mass.

Predict a new observation. Fit the frequency evolution in part of an outburst and predict the next state’s centroid, width, energy dependence, and phase-resolved spectrum. A flexible radius per observation is interpolation, not a mechanism test.

Use independent geometry. Compare the inferred flow radius and inclination with reflection, reverberation, or polarization without multiplying posteriors that share the same spectral state or calibration. Inconsistency is diagnostically valuable.

A truthful reproduction publishes the exact RXTE ObsID and detector list, good-time and state cuts, channel-to-energy mapping by gain epoch, time bin and segment length, dead-time and background treatment, normalization, frequency rebinning, continuum-plus-Lorentzian model, search range, trial accounting, software and CALDB versions, and posterior predictive simulations. Statistical peak uncertainty must remain distinct from the much larger systematic uncertainty in the physical mode identification.

Exercises

For $M=10M_\odot$, $a_*=0.30$, and $r=6$, use $\nu_0\simeq3230\ {\rm Hz}$ to estimate $\nu_\phi$, $\nu_{\rm per}$, and $\nu_{\rm nod}$. An observation contains peaks near $215$, $130$, and $8\ {\rm Hz}$. Does numerical agreement alone establish the relativistic precession model?

Solution

Here $r^{3/2}=14.697$, so

$$\nu_\phi=\frac{3230}{14.697+0.30}\ {\rm Hz}\simeq215.4\ {\rm Hz}.$$

The radial factor is $1-6/r+8a_*/r^{3/2}-3a_*^2/r^2\simeq0.1558$, giving $\nu_r\simeq85.0\ {\rm Hz}$ and $\nu_{\rm per}\simeq130.4\ {\rm Hz}$. The vertical factor is about $0.9258$, giving $\nu_\theta\simeq207.3\ {\rm Hz}$ and $\nu_{\rm nod}\simeq8.1\ {\rm Hz}$.

The match is necessary for this assignment but not sufficient. It was obtained after choosing a common radius and labeling the peaks. A discriminating test must include the search trials, show that the three features are simultaneous, predict their widths and energy-dependent phases, and check the mass or radius independently. A torus-mode model or a different harmonic assignment may fit the centroids with different physical parameters.

References

• HEASARC. n.d. “HEASoft: A Unified Release of the FTOOLS and X-Ray Analysis Software Packages,” version 6.36. Official software page.
• Ingram, Adam, and Sara Motta. 2019. “A Review of Quasi-Periodic Oscillations from Black Hole X-Ray Binaries: Observation and Theory.” New Astronomy Reviews 85:101524. DOI.
• Motta, Sara E., et al. 2014. “Precise Mass and Spin Measurements for a Stellar-Mass Black Hole through X-Ray Timing: The Case of GRO J1655-40.” Monthly Notices of the Royal Astronomical Society 437:2554–2565. DOI.
• Reynolds, Christopher S. 2021. “Observational Constraints on Black Hole Spin.” Annual Review of Astronomy and Astrophysics 59:117–154. DOI.