Comparison of X-ray Emission Properties of TDEs and Soft Flares from AGNs

Technical Summary: Comparison of X-ray Emission Properties of TDEs and Soft Flares from AGNs

Problem Statement
Tidal Disruption Events (TDEs) are rare phenomena where a star is disrupted by a supermassive black hole (SMBH), typically characterized by a sharp rise in luminosity followed by a power-law decay and a soft thermal X-ray spectrum. However, identifying clean TDE samples is challenging because Active Galactic Nuclei (AGNs) can mimic these signatures. AGNs exhibit stochastic X-ray variability, and during flaring states, they can produce softer spectra with hardness ratios (HR) similar to those expected from TDEs. This contamination complicates the search for TDEs in X-ray surveys, particularly as future missions like eROSITA and the Einstein Probe aim to probe SMBH demographics. The central problem addressed is quantifying the overlap in X-ray emission properties between flaring AGNs and TDEs and determining if AGN flares can be distinguished or if they significantly contaminate TDE rate estimates.

Methodology
The authors cross-matched publicly available X-ray point source catalogs from XMM-Newton (4XMM-DR14s) and Swift-XRT (LSXPS) with the Million Quasars Catalog (MQC) and optically selected TDEs from the TNS.

• Sample Selection: The study focused on AGNs with at least three observations to identify flare-like events. Flares were defined as events where the count rate exceeded the median by $>2\sigma$ ($\frac{\text{Count} - \text{Median}}{\sigma} > 2$) and the peak hardness ratio was $HR < -0.75$.
• Spectral Analysis: For XMM-Newton sources with sufficient counts ($>1000$), individual spectra were fitted using XSPEC with absorbed power-law and absorbed blackbody-plus-power-law models. For fainter sources, spectral stacking was performed using Xstack, shifting spectra to the common rest-frame. Similar stacking was applied to the larger Swift-XRT sample, categorized by HR and flare strength.
• Counterpart Identification: Optical counterparts were searched for in the Zwicky Transient Facility (ZTF) catalog to test for stochastic variability consistent with a Damped Random Walk (DRW) model. Infrared counterparts were cross-matched with the WISE catalog, applying a color cut of $W1 - W2 > 0.8$ mag to identify AGNs.
• Rate Estimation: Flare rates were calculated per galaxy per year, accounting for the fraction of galaxies hosting AGNs (estimated between 9.3% and 21.2% based on redshift).

Key Results

1. Spectral Overlap: There is a significant overlap in the distribution of HR and count rate variations between AGNs and TDEs. While TDEs cluster near $HR \sim -1$, flaring AGNs frequently exhibit similar softness ($HR < -0.75$) during flare states.
2. Origin of Softening: Spectral modeling reveals that the softening in flaring AGNs is driven by two factors:
   • The emergence of a blackbody-like component below $\sim 2$ keV, likely associated with the soft X-ray excess (modeled as a warm corona or ionized reflection).
   • The steepening of the continuum power-law component as the accretion rate increases.
   • Stacked spectral analysis confirms that the blackbody component is statistically required to fit the data for soft flaring AGNs ($p$-value $\le 4.7 \times 10^{-3}$) and that its contribution increases with flare strength and lower HR.
3. Flare Rates:
   • High-Intensity Soft Flares: 2.5% (23/920) of XMM-Newton AGNs and 4.4% (179/4089) of Swift-XRT AGNs displayed flares with $\text{Count}/\sigma > 2$ and $HR < -0.75$. The estimated rate is $(1.1\text{--}2.5) \times 10^{-3}$ and $(0.9\text{--}2.0) \times 10^{-3}$ galaxy$^{-1}$ yr$^{-1}$, respectively. This is 34–147 times higher than the current optical TDE rate ($3.2 \times 10^{-5}$ galaxy$^{-1}$ yr$^{-1}$).
   • Extreme Flux Changes: When restricting to flares with a maximum flux change ($\text{peak}/\text{min}$) $> 20\times$ over a rest-frame timescale of $\ge 2$ years, the rate drops to $(4.7\text{--}11.0) \times 10^{-5}$ and $(3.1\text{--}7.1) \times 10^{-5}$ galaxy$^{-1}$ yr$^{-1}$ for XMM-Newton and Swift-XRT, respectively. These rates are comparable to or lower than the TDE rate.
4. Counterpart Identification: Multi-wavelength follow-up is effective but distance-dependent.
   • Swift-XRT sources (typically brighter and lower redshift, $z < 0.5$): 79% were confirmed as AGNs via ZTF variability or WISE color.
   • XMM-Newton sources (typically fainter and higher redshift, $z > 1$): Only 61% were confirmed as AGNs.
5. Spectral Differentiation: Comparing specific TDEs (ASASSN-14li, AT2018fyk, AT2024tvd) with soft AGN flares shows that while some TDEs (e.g., ASASSN-14li) are distinct due to pure blackbody dominance, others (e.g., post-peak AT2018fyk) have spectral shapes nearly indistinguishable from flaring AGNs. A modified HR definition using 0.3–1 keV and 1–10 keV bands can separate the softest TDEs from AGNs but risks missing TDEs with harder emission components.

Significance and Claims
The paper claims that AGN flares can mimic TDEs not only in light curve variability but also in X-ray spectral properties, specifically through the emergence of a soft blackbody component and continuum steepening. The authors conclude that relying solely on X-ray rise times or power-law decay in sparsely sampled surveys will lead to significant contamination by AGN flares.

The study suggests that to isolate a clean TDE sample, future searches (e.g., with eROSITA or Einstein Probe) must:

1. Apply stricter criteria, such as requiring a maximum flux change $> 20\times$ over a baseline of two years or more, which reduces the AGN contamination rate to levels comparable to the TDE rate.
2. Utilize multi-wavelength follow-up (optical variability and IR colors) to identify and filter out AGNs, acknowledging that this method is less effective for faint, high-redshift sources.
3. Exercise caution in interpreting X-ray spectral indices ($\Gamma$) or hardness ratios alone, as the spectral properties of flaring AGNs can overlap significantly with those of TDEs, particularly those that are not purely blackbody-dominated.