Lense-Thirring precession after a supermassive black hole disrupts a star

The study investigates whether Lense-Thirring (LT) precession of an accretion disk formed after a tidal disruption event (TDE) can explain quasi-periodic X-ray variability observed from the nucleus of a quiescent galaxy hosting a supermassive black hole (SMBH). Theory predicts that a newly formed, initially misaligned disk around a spinning SMBH will experience relativistic frame-dragging torques, leading to disk precession that should cease as the disk aligns with the SMBH spin. The TDE AT2020ocn (ZTF18aakelin) at redshift z = 0.0705 provides an opportunity to test these predictions with dense X-ray monitoring. The key questions are: (1) do the observed soft X-ray light curves and spectral temperatures exhibit coherent, repeating modulations consistent with LT precession; (2) are similar modulations absent at optical–UV wavelengths as expected for inner-disk–origin X-rays; and (3) can the observed quasi-periodicity be used to constrain the SMBH spin given reasonable TDE and disk parameters. Establishing LT precession in TDEs would provide a new avenue to measure SMBH spins and probe relativistic accretion dynamics.

Prior theoretical work predicted observable LT precession signatures in TDE disks and their duration before alignment (e.g., Stone & Loeb 2012; Franchini et al. 2016). Radiation-pressure instability (RPI) has been invoked to explain quasi-periodic outbursts in stellar-mass black hole X-ray binaries such as GRS 1915+105 and IGR J17091–3624 (Lightman & Eardley 1974; Janiuk et al. 2000), although LT precession of inner disk regions has also been considered for some variability patterns. Disk tearing and precession of discrete annuli are possible in thin disks with misalignment and low accretion rates (Nixon et al. 2012; Raj & Nixon 2021). X-ray quasi-periodic oscillations in stellar-mass systems commonly show coherences below ~10 (McClintock & Remillard 2006). The present work builds on these by testing LT precession against competing scenarios (repeating partial TDEs, debris stream self-interactions, column density changes, and RPI) with high-cadence, multi-band observations of a TDE.

- Target and host: AT2020ocn/ZTF18aakelin, an optical transient from the center of a quiescent galaxy at z = 0.0705 (luminosity distance ~330 Mpc). Host stellar velocity dispersion measured from a pre-outburst SDSS spectrum (12 Feb 2008) using pPXF with MILES templates: σ* = 82 ± 4 km s⁻¹, implying log10(MBH/M⊙) = 6.4 ± 0.6 via the M–σ relation (measurement plus systematic uncertainties). - X-ray observations: High-cadence NICER monitoring began MJD 59041 (11 July 2020), multiple visits per day, focusing on the first ~130 days when flares are prominent. Additional X-ray data from Swift/XRT (pre- and post-discovery, lower cadence) and two early XMM-Newton EPIC-pn observations (18 and 21 July 2020). Analysis mainly in the 0.3–1.0 keV band where source is typically above background. - NICER data reduction: Level-1 data processed with nicerl2 and default screening; hot detectors flagged per GTI via 0–0.2 keV count-rate outliers; background modeled with the 3c50 model; applied level-3 filtering appropriate for faint sources. Background-subtracted count rates computed per GTI. Spectra binned to ≥25 counts/bin (ftgrouppha), modeled in XSPEC/PyXspec with χ² statistics. - XMM-Newton: EPIC-pn data reduced with xmmsas v19.1.0; high-background intervals removed; source spectra extracted from annuli (pile-up corrected), background from nearby regions; grouped to minimum 1 count/bin. - Swift/XRT: xrtpipeline reduction; source and background regions defined (47.1″ source aperture; 70–210″ background annulus); grades 0–12 used. Provided pre-outburst upper limits and filled NICER gaps; verified no contaminating sources in NICER FoV. - Optical/UV: ZTF forced photometry in g and r bands; Swift/UVOT photometry with uvotsource using 5″ aperture; host contribution subtracted and Galactic extinction corrections applied. - Variability analysis: Lomb–Scargle periodogram (LSP) of NICER 0.3–1.0 keV light curve computed; noise continuum modeled with bending power-law + constant and power-law + constant. Broad peak detected near 17 days with harmonics; apparent modulation ~15 days in light curve. - Significance estimation: Extensive Monte Carlo simulations to assess global false alarm probability (FAP), accounting for: (1) a range of noise continuum models; (2) search over 1–100 day periods/frequencies; (3) identification of broad peaks with coherences from 2 to 10. Derived global FAP < 10⁻⁴ (>3.9σ assuming Gaussianity). - Time-resolved spectroscopy: NICER spectra over first 130 days analyzed to track thermal components. Found two thermal components (cool and warm). When cool component not required, its temperature fixed at 0.062 keV and flux upper limits evaluated. Tracked temperatures and luminosities of components versus time to test for phase-coherent modulations. - Physical modeling: Considered multiple scenarios—repeating partial TDEs, repeated debris stream self-interactions, variable neutral column density, RPI, and LT precession. RPI timescales matched with outer disk truncated at ~30 Rg but overpredicted amplitudes by >2 orders of magnitude unless damped by inner-disk magnetic field B ~10⁴ G. Adopted a rigid-body LT precession model for a geometrically thick disk (H/R > α) for early post-TDE phases. Compared observed rest-frame period (~15.9 days) with precession period–spin relations across MBH uncertainty and disk density profiles to constrain spin (see Fig. 4).

- Discovery of strong, quasi-periodic soft X-ray (0.3–1.0 keV) flux modulations in AT2020ocn, spaced by roughly 15 days and persisting for ~130 days early in the TDE evolution. Similar modulations are absent in the optical–UV bands. - LSP shows a broad, highly significant peak near 17 days with harmonics; adopting the visually consistent 15-day spacing, the signal is termed a 15-day quasi-periodicity. Global false alarm probability is <10⁻⁴ (>3.9σ) across tested noise models, period search ranges (1–100 days), and coherence values (2–10). - Time-resolved X-ray spectroscopy reveals two thermal components (cool and warm). The overall X-ray flux modulates quasi-periodically, accompanied by temperature modulations on the same timescale. Variability of the cool component broadly tracks the warm component, supporting precession of an extended disk rather than a narrow ring. - Lack of optical–UV modulations is consistent with those bands not tracing direct inner-disk emission, whereas soft X-rays originate from the precessing inner flow. - The modulation lifetime (~130 days) is consistent with theoretical expectations that rigid-body precession in SMBHs of 10⁵–10⁷ M⊙ lasts ~0.4–0.7 years before alignment. - Alternative mechanisms are disfavored by the data: repeating partial TDEs, debris stream self-interactions, and variable neutral column density do not match the observed multi-wavelength and spectral behaviors (per Methods). RPI can reproduce the timescale if the disk is truncated at ~30 Rg, but predicts amplitudes >100× larger than observed unless fine-tuned (e.g., B ~10⁴ G) to damp variability. - Using the rigid-body LT precession model with standard TDE parameters (solar-like star; disk extending to the tidal/circularization radius) and the host’s M–σ-based MBH estimate (log10 MBH/M⊙ = 6.4 ± 0.6), the SMBH spin is constrained to 0.05 ≤ a ≤ 0.5, accounting for uncertainties in disk density profiles and MBH.

The observations provide a coherent picture in which Lense-Thirring precession of a newly formed, geometrically thick accretion disk produces quasi-periodic variations in both soft X-ray flux and thermal temperatures with a rest-frame period of ~15.9 days. The absence of corresponding optical–UV modulations supports an origin close to the SMBH, as optical–UV emission likely arises from larger radii (e.g., stream–stream collisions or reprocessing) that would not be strongly affected by inner-disk orientation changes. The duration of the quasi-periodic phase (~130 days) aligns with expectations that, as the accretion rate declines, the disk transitions out of the rigid-precession regime and aligns with the SMBH spin, terminating the modulations. While RPI can, in principle, match the timescale, it significantly overpredicts variability amplitudes unless multiple parameters are tuned (e.g., strong magnetic damping), making LT precession the simpler explanation consistent with the ensemble of observables. The inferred spin range (0.05–0.5) demonstrates that quasi-periodic X-ray modulations in TDEs can serve as an independent probe of SMBH spin, complementing other methods. High-cadence X-ray monitoring thus reveals ordered relativistic phenomena amid otherwise chaotic TDE light curves.

This study reports the discovery of ~15-day, high-significance quasi-periodic soft X-ray flux and temperature modulations during the early evolution of the TDE AT2020ocn. The multi-wavelength and spectral-timing characteristics support an interpretation as Lense-Thirring rigid-body precession of a misaligned, geometrically thick accretion disk around a SMBH of mass ~10^6.4 M⊙. Under standard TDE and disk assumptions, the precession period yields a SMBH spin constraint of 0.05 ≤ a ≤ 0.5. These results highlight the power of high-cadence X-ray timing of TDEs to uncover relativistic disk dynamics and to measure SMBH spins. Future wide-field surveys (e.g., Rubin Observatory/LSST) combined with rapid X-ray follow-up could discover many TDEs exhibiting similar precession-like modulations, enabling population studies of SMBH spin and accretion physics.

- Physical interpretation relies on assumptions: standard TDE parameters (solar-like star), disk extending to the tidal/circularization radius, and rigid-body precession (geometrically thick disk, H/R > α). Deviations from these could alter the inferred spin. - SMBH mass is derived from the M–σ relation with substantial systematic uncertainty (±0.6 dex), which broadens spin constraints. - Disk density profile and geometry uncertainties impact the precession period–spin relation; results are presented across a range of profiles but remain model-dependent. - Alternative mechanisms (e.g., radiation-pressure instability) cannot be completely ruled out; matching observed amplitudes requires fine-tuning (e.g., strong magnetic damping), but residual degeneracies remain. - The quasi-periodicity is broad (low coherence), and the detection significance depends on the adopted noise model and Monte Carlo procedure, though multiple continuum models were tested. - Analysis focuses on the first ~130 days when modulations are present; later-time behavior and potential transitions to thin-disk or disk-tearing regimes are not fully constrained by the presented data.