Detection of disk-jet co-precession in a tidal disruption event

TL;DR

This work presents the strongest observational case yet for disk-jet co-precession in a tidal disruption event, AT2020afhd, via unprecedented 19.6-day quasi-periodicities seen in both X-ray and radio bands. A disk-jet Lense-Thirring precession model, applied to a super-Eddington, thick accretion disk around a ≈ $10^{6.7} M_\odot$ black hole, reproduces the timing and spectral evolution while yielding a positive spin constraint on the black hole. The analysis combines high-cadence, multiwavelength timing with detailed spectral and SED modeling to infer a Doppler-boosted jet with Γ ≈ 1.2–1.6 and magnetic fields of order 10^3 G, consistent with either Blandford-Znajek or Blandford-Payne jet launching. The results demonstrate the diagnostic power of coordinated X-ray and radio monitoring for probing disk-jet physics in TDEs and lay the groundwork for broader searches for LT precession signatures in future events.

Abstract

Theories and simulations predict that intense spacetime curvature near black holes bends the trajectories of light and matter, driving disk and jet precession under relativistic torques. However, direct observational evidence of disk-jet co-precession remains elusive. Here, we report the most compelling case to date: a tidal disruption event (TDE) exhibiting unprecedented 19.6-day quasi-periodic variations in both X-rays and radio, with X-ray amplitudes exceeding an order of magnitude. The nearly synchronized X-ray and radio variations suggest a shared mechanism regulating the emission regions. We demonstrate that a disk-jet Lense-Thirring precession model successfully reproduces these variations while requiring a low-spin black hole. This study uncovers previously uncharted short-term radio variability in TDEs, highlights the transformative potential of high-cadence radio monitoring, and offers profound insights into disk-jet physics.

Figures (13)

• Figure 1: Temporal evolution of the multiwavelength luminosity of AT2020afhd since its optical re-brightening in 2024 (MJD 60310). (A) The unabsorbed X-ray (0.3--2 keV) luminosity. (B) The radio (5--6 GHz) luminosity. The gray-shaded region represents the period used for calculating the cross-correlation function between X-ray and radio data. (C) The UV and optical luminosities. The UVOT data were corrected for Galactic extinction and had the host contribution subtracted, while the ATLAS data were corrected for extinction. The lightcurves are offset as indicated in the legend for clarity. The green line indicates a powerlaw of $t^{-5/3}$. Uncertainties are quoted at the 1$\sigma$ confidence level.
• Figure 2: Temporal evolution of spectral parameters. (A) The diskbb temperature derived from X-ray spectra. The luminosity evolution is depicted with grey symbols to illustrate its correlation with temperature; the two parameters generally co-evolve, with the temperature reaching its peak alongside luminosity during the first 300 days. (B) The in-band spectral index ($F \propto \nu^\alpha$) derived from the VLA (4--8 GHz) and ATCA (4--10 GHz) data. The stars indicate the VLA observations used for the radio SED modeling (Supplementary \ref{['sec:radio_sed']}). (C) The peak frequency, $\nu_{\rm p}$, and the peak flux, $F_{\rm p}$, derived from the SED modeling.
• Figure 3: Timing analysis of the X-ray and radio data. (A) Lomb-Scargle periodogram of the X-ray lightcurve. The calculation includes data collected between August 3 and October 21, 2025, during which the X-ray quasi-periodic variations were clearly observed. (B) Cross-correlation function between the X-ray and radio data. The data used in this calculation are indicated by the gray-shaded region in Figs. \ref{['fig:lc']}A and B. The histogram represents the distribution obtained from bootstrap simulations, with the red dashed line marking the median of the distribution at $-19.0$ days. (C) Folded X-ray and radio lightcurves with a period of 19.6 days. The radio data were rebinned into 0.1-phase intervals using a weighted mean for clarity. The data used in this calculation are shown in Fig. \ref{['fig:schematics_model']}B.
• Figure 4: Comparison of TDEs with early intense radio detection. The luminosities of the two on-axis jetted TDEs, Swift J1644+57 and AT2022cmc, as well as AT2020afhd, have been rescaled for clarity. The AMI-LA data at 15.5 GHz for Swift J1644+57, ASSASN--14li, AT2019azh and AT2022cmc were adapted from Berger2012, Bright2018, Sfaradi2022, and Rhodes2023, respectively.
• Figure 5: The disk-jet precession model. (A) Schematics of the proposed disk-jet precession model. $\theta_{\rm obs}$ and $\theta_{\rm i}$ represent the viewing angle of the system and the disk/jet precession angle around the black hole axis, respectively. The left and right plots correspond to the phases of the X-ray and radio variations when the luminosity is relatively low and high, respectively. (B) Comparison of the disk-jet precession model (the lower and upper black curves) with X-ray (0.1--2 keV) and radio (5--6 GHz) observations. In the presented model, we adopted a BH mass of $M_\bullet=10^{6.7} M_{\odot}$, a scaleheight ratio of $H/R\sim1$, and an outer disk radius equal to the circularization radius of the debris. As a result, we determined an inclination angle of $\theta_{\rm obs}\sim37.8-38.9^\circ$, a disk/jet precession angle of $\theta_{\rm i} \sim 14-15^\circ$, and a Doppler factor of $\Gamma\sim1.2-1.6$.