Tidal disruption events (TDEs) are usually sought in quiescent galaxies, where a flare stands out against the lack of nuclear activity. But in active galaxies, where the supermassive black hole is already surrounded by an accretion disk and a dusty torus, such events long went unnoticed. Optical and X-ray surveys are blind to flares hidden behind dust, and a large fraction of active nuclei live in exactly such conditions. Earlier work, such as Masterson et al. (2024), found infrared TDEs in quiescent galaxies but did not address active ones. This work is the first to systematically explore extreme mid-infrared variability in a sample of nearly one hundred thousand active nucleus candidates selected by WISE color.
The authors used the WISE AGN R90 catalog, containing 4.5 million reliable active nucleus candidates, and selected a pilot sample of 600,000 objects. For each source, they built light curves in the W1 (3.4 μm) and W2 (4.6 μm) bands from WISE and NEOWISE data spanning 15 years with roughly half-year steps. Instead of flux, they analyzed the change in W1−W2 color, since it is sensitive to dust heating. To assess variability, they computed three metrics: weighted standard deviation, amplitude (the difference between the 95th and 5th percentiles), and reduced chi-square relative to a constant color. They then combined these into a single score and selected objects with extreme values. For the two best candidates, they obtained optical spectra with the 200-inch telescope at Palomar Observatory and compared them with archival DESI data. The flares were modeled as dust echoes from a thin spherical shell, using the canonical mass fallback rate $t^{-5/3}$.
Two sources with extreme color variability were found: J2244+0816 (z=0.1185) and J2150−1132 (z=0.2605). Both showed a brightening of 1.5–2 magnitudes in W1 and W2, followed by a decline below the pre-flare level. For J2244+0816, a simultaneous optical flare was recorded with a peak absolute magnitude of −23.4 and a rapid decline: the brightness halved in 77 days in the rest frame, steeper than the classical $t^{-5/3}$ rate. For J2150−1132, no optical flare is seen, indicating strong dust extinction. Modeling the dust echo gave dust shell radii of 0.14 and 0.19 parsecs, dust temperatures around 1300 and 1170 K, and total radiated energies of $10^{51}$.67 and $10^{51}$.49 erg. Spectroscopy of J2150−1132 showed the disappearance of the broad Hα component between 2021 and 2025, classifying the object as a changing-look active nucleus. Both events lie in the region of the peak luminosity–decay time diagram occupied by extreme nuclear transients and are two orders of magnitude brighter than supernovae.
This work demonstrates for the first time that the WISE infrared survey can find tidal disruption events in active galaxies, including completely dust-hidden events that are invisible in the optical and X-ray. Unlike previous studies focused on quiescent galaxies, it shows that active nuclei also produce powerful flares and that these can be identified by changes in infrared color. The detection of a rapid optical decline in J2244+0816 points to a possible partial disruption of a star or the disruption of a star by an intermediate-mass black hole embedded in the accretion disk. This is the first observational evidence for micro-TDEs, which have been predicted theoretically. Furthermore, the connection between the flare and the subsequent weakening of nuclear activity supports the retrograde TDE model, where ejected material collides with the inner disk and temporarily quenches accretion.
The topic could develop in several directions. First, expanding the sample to all 4.5 million R90 candidates and relaxing quality criteria would allow finding more events and estimating their rate. Second, future infrared surveys with higher cadence, such as the Roman Space Telescope, could trace the fast phases of flares in detail and distinguish between full and partial disruptions. Third, hydrodynamic simulations of TDEs in accretion disks, accounting for the orientation of the stellar orbit, would help interpret the observed diversity of light curves. Finally, a systematic search for micro-TDEs in active galactic nuclei disks could confirm or refute their predicted high rate.
The results will affect several areas of astrophysics: the study of tidal disruption events, the physics of active galactic nuclei, the dynamics of stellar clusters near supermassive black holes, and the modeling of dusty tori. They are also important for planning future infrared time-domain surveys.
Immediate next steps include expanding the search to the full R90 sample, obtaining additional spectra to confirm the nature of the events, and more detailed modeling of the light curves that accounts for the geometry of the dusty torus. X-ray observations of J2150−1132 are also planned to test the hypothesis of a hidden accretion disk.
This work directly addresses the unresolved problem of the rate of tidal disruptions in active galaxies and their contribution to the growth of supermassive black holes. It also touches on the existence of micro-TDEs and the role of intermediate-mass black holes in accretion disks. Moreover, the discovery of a changing-look active nucleus after the flare sheds light on the mechanisms that switch nuclear activity on and off, which remains one of the mysteries of active galaxy physics.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
The authors carried out a systematic search for extreme mid-infrared variability among roughly 600,000 active galactic nucleus candidates from the WISE AGN R90 catalog, analyzing the W1−W2 color over 15 years. They found two bright flares (J2244+0816 and J2150−1132), interpreted as tidal disruption events, one of which is invisible in the optical due to dust extinction. Modeling the dust echo yielded dust shell radii of ~0.14–0.19 pc, temperatures of ~1170–1300 K, and total energies of ~10^51.5–10^51.7 erg.
Where its strength is
The work demonstrates that the mid-infrared range can uncover TDEs in active galaxies that are hidden from optical and X-ray surveys by dust, and it paves the way for a complete census of such events.
Where it could go next
Worth trying
Why this may matter
A complete census of TDEs in active galaxies is important for understanding the population of supermassive black holes and their environments, as well as for assessing the contribution of such events to black hole growth and the enrichment of the interstellar medium.
Nearby works in our archive
The archive is sparse around this work: the closest review is only 0.56, against a usual 0.64. It looks like few have drilled nearby.
The flare in J2150−1132 was so hidden by dust that it was completely invisible in the optical, even though it brightened by two magnitudes in the infrared. It is like a powerful explosion happening behind a thick curtain: we only see the curtain heating up, not the explosion itself.
The micro-TDE scenario, where a star is torn apart by a small black hole hiding in the accretion disk of a giant black hole, echoes science fiction ideas about secondary black holes populating the vicinity of supermassive objects, such as in Terry Goodkind's novel 'Pillars of Creation' or in some episodes of 'Star Trek' where characters encounter unexpected sources of gravity in dense stellar clusters.