Utilizing Maximum Variability to Discern TDE Emission from AGN Flares

TL;DR

The study tackles the difficulty of distinguishing TDE X-ray flares from AGN variability in wide-field surveys by quantifying AGN X-ray variability with ensemble structure functions (SF) and a time-dependent MaxVar metric. Using cross-matched XMM-Newton and Swift-XRT data with the MILLIQUAS catalog, the authors build sizable AGN samples and a catalog of TDEs, then compute SFs and MaxVar to compare variability between TDEs and AGNs. They find AGN SF follows a shallow power law with γ≈0.11–0.14 and exhibits long-timescale turnover with a damping timescale of $τ≈950±300$ days, while TDE-like variability is prominent at short lags (≤20 days) but becomes less frequent at longer lags in AGN data. The results imply that many eROSITA TDE candidates could be consistent with AGN flares, underscoring the need for variability-based discrimination in upcoming all-sky X-ray surveys and guiding contamination estimates for TDE searches.

Abstract

X-ray emission arising from active galactic nucleus (AGN) activity may potentially mimic the expected emission of tidal disruption events (TDEs). Ongoing and upcoming wide-field X-ray surveys will detect thousands of TDE-like sources, and classifying them securely as TDEs or AGNs is a challenging task. To this aim, we measure the average X-ray variability of AGNs and derive a threshold of maximum variation as a function of time separating the TDEs from AGN flares. For the comparison between TDE and AGN X-ray variability, we cross-match the publicly available XMM-Newton and Swift-XRT point source catalogs with the Million Quasars Catalog and optically selected TDEs. Then we compute the X-ray structure function (SF) and maximum variability of the AGN and TDE samples. The X-ray SF of AGNs has a power-law index $γ\sim0.11-0.14$ when fitted with a simple power-law model. However, the SF of AGNs is best described by a broken power-law or a power exponential model with a damping time scale $τ=950\pm300$ days. The maximum variability comparison between TDE and simulated AGN light curves indicates they have a similar order of variation on a time scale of less than 20 days. However, at a longer time scale of $\sim20$ days or more, the large-scale variations expected from power-law-like decay in TDEs become less frequent in AGNs. Furthermore, we compare the maximum variability of eROSITA TDE candidates with AGN, finding that many of the eROSITA-DE TDE candidates are consistent with flares from AGNs, and may not be TDEs.

Figures (3)

• Figure 1: The nearest neighbor distance distribution obtained by comparing the XMM-Newton (top panel) and Swift-XRT (bottom panel) sources with MILLIQUAS catalog. The histograms show a bimodal distribution, which is modeled with two Gaussians. We use these fits to obtain the maximum distance used for cross-matching the X-ray sources with the AGN sample. The black dashed line shows the 3$\sigma$ distance from the mean of the nearest Gaussian.
• Figure 2: The illustration of MaxVar from an artificial light curve that consists of six measurements.
• Figure 3: SF (in log-log scale) of AGN sample computed from the XMM-Newton and Swift-XRT data sets. The continuous lines indicate a simple power-law model fit to the data. The dashed lines indicate a slightly different power-law index obtained in a similar study by vagnetti2011 using XMM-Newton and Swift-XRT data sets.