QPO Properties in Accretion Systems

• QPO properties are defined as quasi-periodic oscillations seen as narrow peaks in the power spectrum, characterized by centroid frequency, quality factor, and amplitude.
• They provide insights into disk-corona geometry, relativistic effects, and inner accretion flow dynamics across black hole binaries, neutron star systems, and AGN.
• Spectral-timing analyses and time-dependent Comptonization models constrain mechanisms like Lense–Thirring precession and heartbeat oscillations in diverse accretion states.

Quasi-Periodic Oscillation (QPO) Properties

Quasi-periodic oscillations (QPOs) are coherent or semi-coherent features observed in the time series of accreting astrophysical sources, manifested as narrow peaks superimposed on broadband noise in the Fourier power spectral density (PSD). QPOs are prominent in the X-ray emission of accreting black hole X-ray binaries, neutron star low-mass X-ray binaries, cataclysmic variables, active galactic nuclei (AGN), and gamma-ray bursts (GRBs), with characteristic frequencies tracing the dynamical, thermal, and radiative phenomena in the innermost regions of accretion flows. Their rich phenomenology—centroid frequency, quality factor, amplitude, spectral-timing behavior—encodes information about the structure, geometry, and relativistic dynamics in extreme gravitational environments.

1. Taxonomy, Frequency Ranges, and Harmonic Structure

QPOs are classified by their centroid frequency ν₀, coherence Q ≡ ν₀/Δν (where Δν is the FWHM), and fractional rms amplitude. In black hole X-ray binaries, several distinct QPO types are identified (Motta, 2016, Ingram et al., 2020, Alston et al., 2015):

• Low-Frequency QPOs (LFQPOs, ν₀ ≲ 50 Hz):
  • Type-C: ν₀ ≈ 0.1–30 Hz; strong (rms ≲ 20%), narrow (Q ≳ 10), present with flat-top broadband noise; often exhibits subharmonics and harmonics (Pawar et al., 2015, Rao et al., 2010).
  • Type-B: ν₀ ≈ 1–6 Hz; moderate amplitude (rms ≈ 2–5%), Q ≳ 6, appears in soft-intermediate states.
  • Type-A: ν₀ ≈ 6–8 Hz; weak (rms ≲ 2%), broad (Q ≲ 3); associated with high-soft state (Zhang et al., 2023).
• High-Frequency QPOs (HFQPOs, ν₀ ≳ 60 Hz): Observed at 60–450 Hz, typically Q ≈ 5–15, low amplitude (rms ≈ 0.5–6%), sometimes in commensurate pairs (3:2 frequency ratio) (Motta, 2016, Alston et al., 2015).

In neutron star LMXBs, kHz QPOs are detected at ν₀ ≈ 500–1200 Hz, with both a "lower" and "upper" QPO often present simultaneously, differing in spectral-timing signatures (Troyer et al., 2018, Ribeiro et al., 2017).

In AGN, relatively narrow (Q ≃ 8–12), low-amplitude (rms ≈ 5–6%) QPOs have been observed at ν₀ ≈ 2 × 10⁻⁴ Hz (1–2 hr period); properties and scaling relations suggest direct analogy with HFQPOs in XRBs (Alston et al., 2015, Li et al., 1 Feb 2026).

2. Measured Metrics: Frequency, Quality Factor, Amplitude, and Lifetime

Centroid frequency ν₀ and FWHM Δν are obtained via Lorentzian fits to the PSD:

• QPOs show a broad distribution of Q, from Q ≈ 2–3 (type-A QPOs) to Q ≳ 10 in typical type-C QPOs (Zhang et al., 2023, Motta, 2016).
• Exceptionally, some highly-coherent QPOs in "heartbeat" black hole binaries reach Q ≳ 50–60 at ν₀ ≈ 5–8 Hz (Wang et al., 2024).

Fractional rms amplitude is defined as the square root of the integrated Lorentzian power, usually in fractional rms² units (Motta, 2016, Alston et al., 2015). For example:

• Type-C QPOs: rms ≈ 5–20%
• Type-A QPOs: rms ≈ 0.9% at 7 Hz in MAXI J1348–630 (Zhang et al., 2023)
• HFQPOs: rms ≲ 6%
• kHz QPOs in neutron stars: fractional rms increases with photon energy, reaching ≈15–20% by 12 keV (Troyer et al., 2018).

Lifetime and intermittency:

• QPO signals are often intermittent, especially in black-hole binaries during the hard state; in MAXI J1820+070, low-frequency type-C QPOs occur in trains lasting ∼5 cycles (Q≈4–6), as revealed by wavelet Z-transforms (Zhang et al., 2023).

Harmonic and subharmonic content:

• Type-C QPOs exhibit a robust 1:2:3:4 (sometimes up to 1:2:3:4:6) harmonic sequence, with the fundamental and first harmonic Q ≈ constant (frequency modulation signature), the subharmonic showing almost constant width (amplitude modulation) (Pawar et al., 2015, Rao et al., 2010).

3. Energy Dependence: rms and Phase-Lag Spectra

Fractional rms vs. energy:

• LFQPO amplitudes nearly always rise with photon energy, often following rms(E) ∝ E^γ, with γ in the range 0.2–0.8 depending on spectral state; above a certain threshold (e.g., ≈15 keV), the rms plateaus (Xu et al., 26 Jun 2025, Alston et al., 2015, Troyer et al., 2018).
• For type-A QPOs, rms rises from <1% at 1 keV to 3% at 6 keV, tracking the Comptonised component (Zhang et al., 2023).
• Similar steep (and saturating) energy dependence is found for kHz QPOs (neutron stars), AGN QPOs, and X-ray QPOs in cataclysmic variables (Troyer et al., 2018, Alston et al., 2015, Smak, 2014).

Phase/time-lag spectra:

• Spectral-timing analysis (cross-spectrum phase lag Δφ(E) or time lag Δτ(E) = Δφ(E) / 2πν₀) reveals both "hard" (higher energy photons lag) and "soft" (higher energy photons lead) lag modes.
  • Type-A QPOs: persistent soft lags (Zhang et al., 2023).
  • Type-C QPOs: lag sign and magnitude depend on inclination and frequency (Ingram et al., 2020, Li et al., 1 Feb 2026).
  • In neutron star systems, lower kHz QPOs show "soft lags" (i.e., high-energy photons lead by tens of μs), while upper QPOs have negligible lags (Troyer et al., 2018).
  • AGN QPOs: soft X-ray bands and Fe Kα emission lag the continuum by hundreds to ∼1000 s, interpreted as reverberation lags from reflection/reprocessing (Alston et al., 2015, Li et al., 1 Feb 2026).
• The lag–energy spectra at the QPO frequency encode the size and geometry of the Comptonizing region and any feedback to the disk (Zhang et al., 2023, Rout et al., 2023, Troyer et al., 2018).

4. Correlations with Spectral State and Physical Interpretation

Spectral state dependence:

• QPO type and properties (frequency, amplitude, lag) are tightly coupled to spectral state.
  • Type-C QPOs dominate in the hard/hard-intermediate state, with increasing ν₀ as the spectrum softens and the disk fraction increases (Motta, 2016, Rao et al., 2010).
  • Type-B and -A QPOs emerge in transition and high-soft states, respectively (Zhang et al., 2023, Motta, 2016).
  • Extremely coherent QPOs (Q > 50) are found in disk-dominated "heartbeat" states with very low broadband rms (Wang et al., 2024).

Empirical correlations:

• ν_QPO increases with disk flux/temperature, although the relation can be non-monotonic with "double-branch" behavior: negative correlation at ν_QPO ≲ ν_tr, positive above (Xu et al., 26 Jun 2025).
• kHz QPO frequencies in neutron stars scale identically with coronal parameters (Γ, kT_e, τ) after applying a constant offset between lower and upper QPO, suggesting a common dynamical origin (e.g., Keplerian/orbital or resonance) but distinct radiative mechanisms for amplitude (Ribeiro et al., 2017).

Physical models and Comptonization geometry:

• The amplitude and lag spectra favor a Comptonized origin for QPO modulation, with the hot corona or boundary layer as the oscillating region (Troyer et al., 2018, Zhang et al., 2023, Rout et al., 2023).
• Model fits with time-dependent Kompaneets (e.g., vkompthdk) constrain coronal size, feedback, and geometry:
  • Type-A QPOs: vertically-extended, partially disk-covering corona, L ≈ 2300 km (Zhang et al., 2023).
  • Jointly-present type-B/type-C QPOs in GRO J1655–40: slab-like, disk-covering corona (type-C) vs. compact, jet-like corona (type-B), differing in feedback efficiency and size (Rout et al., 2023).

Modulation mechanism:

• Frequency modulation is the dominant process for the type-C fundamental and harmonic (constant Q), while amplitude modulation is important for subharmonics (Pawar et al., 2015, Rao et al., 2010).

5. Theoretical Frameworks: Disk and Jet Precession, Relativistic Effects

Lense–Thirring precession:

• The most successful framework for type-C QPOs is geometric Lense–Thirring precession of an inner hot flow (or precessing jet/corona) misaligned with the black hole spin (Ingram et al., 2020, Zhang et al., 2023, Xu et al., 26 Jun 2025).
• Precession frequency:

$$\nu_{\mathrm{LT}}(r) = \frac{1}{2\pi} \frac{c^3}{GM} a_{\ast} r^{-3}$$

where $a_{\ast}$ is the dimensionless spin and $r=R/R_g$ (Xu et al., 26 Jun 2025).

• QPO properties reflect the co-evolution of disk truncation radius, hot flow, and covering fraction; state transitions manifest as changes in these geometric and physical parameters (Xu et al., 26 Jun 2025, Zhang et al., 2023).

Spectral-timing diagnostics:

• Energy-dependent lags and amplitude spectra are consistent with the timing signature expected from precession or oscillations in the corona/inner flow, supported by frequency–dependent reflection reverberation and iron-line tomography (Zhang et al., 2023, Alston et al., 2015, Rout et al., 2023).

High-coherence QPOs associated with "heartbeat" states:

• Oscillations with Q ≳ 50–60 in IGR J17091–3624 and GRS 1915+105 are likely tied to inner disk limit-cycle instabilities coupled to the corona or jet base, rather than standard precession models (Wang et al., 2024).

QPOs in non-XRB systems:

• AGN QPOs—by frequency scaling and phenomenology—are likely analogues of XRB HFQPOs, with lags and rms diagnostic of disk-corona reverberation (Alston et al., 2015, Li et al., 1 Feb 2026).
• In cataclysmic variables and white dwarf systems, QPOs typically have lower coherence, shorter lifetimes, and frequency drifts on hour timescales, attributed to transient disk oscillations or boundary-layer instabilities (Smak, 2014).

6. Rms–Flux Relation, Filtering, and Population Distinctions

rms–flux relation:

• Type-C QPOs show a linear positive rms–flux correlation at low ν₀, flattening and then turning negative as ν₀ increases above ≈6 Hz (Heil et al., 2010). This is attributed to a physical or radiative "filter" (suppression of high-frequency variability), which, once modeled, restores a flat (saturated) intrinsic rms–flux relation at high ν₀.

Quality factor and classification distinctions:

• Q ≲ 3 (Type-A, soft state); Q ≳ 10 (Type-C, hard/intermediate); Q > 50 (heartbeat state, not explained by existing type A/B/C schema) (Motta, 2016, Zhang et al., 2023, Wang et al., 2024).
• The relative prevalence of amplitude and frequency modulation, energy dependence, phase-lag sign, and coupling to reflection all serve to differentiate QPO sub-classes and associate them with underlying geometries and physical modes.

Table: Key QPO Properties Across Classes

QPO Type	Typical ν₀ (Hz)	Q	rms (%)	Spectral State	Lag Sign
Type-C (BH)	0.1 – 30	8–15	5–20	Hard/Intermediate	Hard or soft
Type-B (BH)	1 – 6	6–12	2–5	Soft-Intermediate	Distinct
Type-A (BH)	6 – 8	1–3	≤2	High-Soft	Soft
HFQPO (BH)	60–450	5–20	0.5–6	High-flux Intermediate/Soft	Hard/soft
kHz QPO (NS)	500–1200	5–20	Up to 20	Atoll/Z source, luminous	Soft/zero
AGN QPO	2e–4 Hz	8–12	5–6	Hard continuum band	Soft
Heartbeat QPO	5–8	50–60	2–5	Disk-dominated/Heartbeat	Steep/soft

7. Outstanding Issues and Future Constraints

Key uncertainties include the physical origin of the subharmonic QPOs, detailed mechanisms for amplitude vs. frequency modulation, direct mapping between lag properties and coronal geometry, and the precise triggers of high-coherence QPOs unique to "heartbeat" states. Technological advances in phase-resolved spectroscopy, X-ray polarimetry, and high count-rate observations are poised to transform the capacity of QPO observations to probe strong gravity and plasma physics (Ingram et al., 2020).

Spectral-timing studies, including joint rms and lag-spectra modeling with time-dependent Comptonization codes (vkompthdk), are establishing a model-independent basis for inferences on coronal feedback, disk-corona coupling, and resonant amplification phenomena (Zhang et al., 2023, Rout et al., 2023). Systematic multiwavelength and polarimetric diagnostics will further refine the connection between relativistic precession, inner flow structure, and broadband variability.

---

References: (Pawar et al., 2015, Rao et al., 2010, Alston et al., 2015, Motta, 2016, Zhang et al., 2023, Zhang et al., 2023, Troyer et al., 2018, Ribeiro et al., 2017, Heil et al., 2010, Rout et al., 2023, Wang et al., 2024, Zhang et al., 2023, Li et al., 1 Feb 2026, Ingram et al., 2020, Xu et al., 26 Jun 2025, Alam et al., 2014, Smak, 2014).