Ansky: Extreme QPE and TDE-like Transient

• Ansky is a nuclear transient source characterized by extreme X-ray quasi-periodic eruptions (QPEs) with delayed UV flares and evolving absorption features.
• Its X-ray eruptions have doubled in recurrence time and duration from 2024 to 2025, with a positive period derivative indicating non-standard dynamical processes.
• Multi-wavelength observations, including featureless UV spectra and detailed timing analysis, position Ansky at the intersection of QPE, tidal disruption event (TDE), and low-level AGN phenomenology.

Ansky, also designated ZTF19acnskyy and associated with the nucleus of SDSS1335+0728, is a transient nuclear source that brightened abruptly in December 2019 and later became the host of extreme X-ray quasi-periodic eruptions (QPEs). It is distinguished by an unusual combination of observables: energetic and long-lived soft X-ray eruptions, rapidly evolving X-ray absorption-like features, a smoothly increasing recurrence period with a directly measured positive period derivative, the first robust delayed ultraviolet (UV) counterpart to X-ray QPEs, and a featureless UV spectrum that resembles a slowly evolving tidal disruption event (TDE) more closely than a standard active galactic nucleus (AGN). Its physical origin remains debated, and current interpretations span EMRI-related orbiter–disk or stream–disk interactions, mass-transfer scenarios, accretion-disk instabilities, and TDE-linked channels (Chakraborty et al., 9 Apr 2025, Hernández-García et al., 19 Sep 2025, Chakraborty et al., 18 Feb 2026, Guo et al., 3 Mar 2026, Zhu et al., 25 Oct 2025, Sánchez-Sáez et al., 1 Jul 2026).

1. Source identity and astrophysical setting

Ansky is a nuclear transient in SDSS1335+0728, observed at $z=0.024$, and associated with a central black hole of order $\sim 10^6\,M_\odot$. The source emerged as an extreme QPE system after its 2019 optical brightening and has since been characterized by recurrence times, energetics, and X-ray spectral structure that surpass most known QPE sources. The system has been discussed in the context of both TDE and AGN phenomenology, but the data do not support a simple classification as either an ordinary turn-on AGN or an ordinary optically selected TDE (Chakraborty et al., 9 Apr 2025, Hernández-García et al., 19 Sep 2025, Sánchez-Sáez et al., 1 Jul 2026).

Several papers treat Ansky as a boundary case linking multiple transient classes. The UV spectral evidence has been interpreted as favoring a featureless, slowly evolving TDE, while the spatially resolved optical and mid-infrared data indicate that the nucleus was not dormant before the 2019 brightening. This suggests that the event occurred in a system with already ongoing low-level accretion, whether as a persisting or gradually fading low-luminosity AGN or as a long-lived TDE remnant disc (Zhu et al., 25 Oct 2025, Sánchez-Sáez et al., 1 Jul 2026).

2. X-ray QPE phenomenology and secular timing evolution

The X-ray phenomenology of Ansky is extreme even within the small QPE population. Its eruptions are long in duration, luminous, and strongly variable, with blackbody-dominated spectra and peak-to-quiescence amplitudes reported as $>500\times$. Phase-resolved analyses describe substantial temperature evolution during the eruptions, and 2025 NICER monitoring showed that the eruptions had become both longer and more energetic than in 2024 (Chakraborty et al., 9 Apr 2025, Hernández-García et al., 19 Sep 2025).

Quantity	2024	2025
Recurrence time	$436.8 \pm 80.3$ ks ($\sim 4.5$ days)	$907 \pm 25$ ks ($\sim 10.5$ days)
Flare duration	$129 \pm 43$ ks ($\sim 1.5$ days)	$243 \pm 35$ ks ($\sim 10^6\,M_\odot$0 days)
Mean integrated eruption energy	4× lower than 2025	$\sim 10^6\,M_\odot$1
Rise time	$\sim 10^6\,M_\odot$2 days	$\sim 10^6\,M_\odot$3 days
Decay time	$\sim 10^6\,M_\odot$4 days	$\sim 10^6\,M_\odot$5 days

The 2025 eruptions were therefore approximately twice as long in both recurrence time and duration, four times more energetic, and more asymmetric because the decay became much longer while the rise remained similar. NICER analyses further reported 14 observed QPEs between January and June 2025 and noted that the recurrence time was increasing at $\sim 10^6\,M_\odot$6 days per flare, consistent with a secular drift (Hernández-García et al., 19 Sep 2025).

A subsequent timing study reported the first direct measurement of a QPE period derivative, finding a smoothly increasing period with

$\sim 10^6\,M_\odot$7

Using 23 eruptions monitored between 2025 and 2026, O–C diagrams and timing models indicated a positive and approximately constant $\sim 10^6\,M_\odot$8 over more than one year, with an initial recurrence period in 2025 of $\sim 10^6\,M_\odot$9 days and later values reaching roughly 14 days. In most EMRI-based pictures this sign of $>500\times$0 is surprising, because orbital periods are usually expected to shrink under gravitational-wave-driven evolution (Chakraborty et al., 18 Feb 2026).

3. Time-resolved X-ray spectroscopy and expanding-debris interpretations

Ansky is the first QPE system in which rapidly varying absorption-like X-ray features have been observed within individual eruptions. These residuals occur in the Wien tail of the continuum, around 0.6–1.2 keV, and evolve dramatically with QPE phase. The centroid energy typically rises from $>500\times$1–0.8 keV before the peak to $>500\times$2–$>500\times$3 keV near maximum, then decreases to $>500\times$4–0.7 keV during decline. Their strengths and widths co-evolve with continuum blackbody temperature and luminosity, and their detection across NICER, XMM-pn, MOS1/2, and RGS argues against instrumental origin (Chakraborty et al., 9 Apr 2025).

Physical modeling with ionized outflow prescriptions yielded column densities $>500\times$5–$>500\times$6, ionization parameters $>500\times$7–4.18, or electron temperatures $>500\times$8–397 eV in collisional fits, together with blueshifted velocities from $>500\times$9 up to 0.4. High-resolution RGS fits also indicated ionized blueshifted emission with bulk velocity $436.8 \pm 80.3$0, emitter column density $436.8 \pm 80.3$1, ionization $436.8 \pm 80.3$2, and a nitrogen over-abundance of $436.8 \pm 80.3$3 times solar. The latter has been compared to the nitrogen enrichment reported in other QPEs and some TDEs (Chakraborty et al., 9 Apr 2025).

The proposed physical picture is an EMRI-like orbiter–disk collision in which each impact launches shock-heated debris that undergoes radiation pressure-driven homologous expansion. In that framework, the observed absorption-line blueshifts trace the expansion velocity of the debris rather than a conventional disk wind. An analytical toy model with $436.8 \pm 80.3$4, $436.8 \pm 80.3$5, and $436.8 \pm 80.3$6 erg per burst was reported to reproduce the absorption phenomenology qualitatively, including P Cygni-like line profiles in a spherical expansion geometry. A large ejected mass is required to keep the ionization low enough for observable X-ray absorption, and the rapid evolution of the line centroid is associated with changes in photospheric velocity and ionization as the debris expands (Chakraborty et al., 9 Apr 2025).

4. Ultraviolet behavior: delayed QPE response and featureless continuum

Ansky is the first QPE source with a clear, recurrent UV emission temporally coupled to the X-ray QPE signal. Intensive, high-cadence Swift monitoring over five consecutive cycles showed periodic brightenings in the UVW2 band coincident with the X-ray eruptions. These UV flares are broader and less impulsive than the X-ray events, with variability amplitude of approximately 30%, far smaller than the $436.8 \pm 80.3$7 amplitude in X-rays. Cross-correlation analysis found a centroid lag

$436.8 \pm 80.3$8

a peak lag $436.8 \pm 80.3$9 days, and $\sim 4.5$0 (exactly 0.56), with a reported $\sim 4.5$1-value of $\sim 4.5$2. The lag is negligible under redshift correction at $\sim 4.5$3, and the UV counterpart remains unique among more than ten known QPE sources (Guo et al., 3 Mar 2026).

Two main physical interpretations were advanced for this delay. In star–disk collision models, the lag can correspond to the diffusion timescale through an optically thick, shock-heated blob; with $\sim 4.5$4 day and $\sim 4.5$5, the timescale and UV flux were reported to match the observations. In mass-transfer or disk-instability models, the lag can instead be a light-crossing delay caused by reprocessing of X-rays emitted near the black hole in an outer UV-emitting region, with $\sim 4.5$6 again yielding a delay of order one day. The viscous timescale is orders of magnitude too long, and the thermal timescale was judged less likely. A central empirical point is that Ansky’s recurrence period is much longer than the measured lag, which reduces temporal blending and may explain why a UV counterpart is detectable here but not in faster QPE systems (Guo et al., 3 Mar 2026).

Independent UV spectroscopy strengthens the interpretation of Ansky as atypical among nuclear transients. HST/STIS observations on 2025 May 24 covered 1100–3050 Å in the rest frame at median S/N of 12 and revealed a featureless, steep continuum fit by

$\sim 4.5$7

or equivalently by a blackbody with $\sim 4.5$8. No broad optical or UV emission lines were detected, even six years after outburst; only narrow absorption in Ly$\sim 4.5$9, NV, and CIV with FWHM $907 \pm 25$0 was attributed to host-galaxy gas. The peak blackbody luminosity was measured as $907 \pm 25$1, at the low-luminosity end of optically selected TDEs, with a rise to peak of about 350 days, $907 \pm 25$2 days, and $907 \pm 25$3 days. These properties were interpreted as more consistent with a slowly evolving featureless TDE than with a standard AGN or weak-lined quasar, and HST astrometry placed the source at the galactic center with offset $907 \pm 25$4 pc and a $907 \pm 25$5 upper limit of 54 pc (Zhu et al., 25 Oct 2025).

5. Host galaxy structure and long-term nuclear accretion history

Spatially resolved optical and mid-infrared spectroscopy has shown that the Ansky event occurred in a dynamically and spectroscopically structured host. VLT/MUSE data revealed two counter-rotating stellar regions and kinematically cold gas with $907 \pm 25$6, a configuration interpreted as consistent with a past minor merger. Stellar-population analysis showed an old host galaxy together with ongoing star formation confined to a ring at intermediate radii (Sánchez-Sáez et al., 1 Jul 2026).

Emission-line diagnostics further indicated a three-zone structure. The central region is powered by SMBH accretion and contains compact high-ionization coronal emission, including [NeVI]$907 \pm 25$7, [NeV]$907 \pm 25$8, and [OIV]$907 \pm 25$9. At intermediate radii, line ratios correspond to a star-forming ring, while the outer galaxy shows LINER-like excitation. In the mid-infrared, low-ionization lines such as [NeII] $\sim 10.5$0 and [SIII] $\sim 10.5$1 are spatially extended and correlate with the star-forming ring, whereas the high-ionization coronal lines remain nuclear (Sánchez-Sáez et al., 1 Jul 2026).

A Balmer-line light-echo analysis was used to reconstruct the nuclear ionization history. It yielded a minimum ionizing luminosity of $\sim 10.5$2 sustained for at least $\sim 10.5$3 years, implying that the nucleus was not dormant before the 2019 optical brightening. JWST/MIRI/MRS spectra also showed strong broad silicate emission at 9.7 and $\sim 10.5$4, interpreted as optically thin dust and therefore inconsistent with a classical AGN dusty torus. The pre-2019 accretion state was accordingly modeled in two ways: a persistent or gradually fading low-luminosity AGN with $\sim 10.5$5, or a long-lived TDE remnant disc. In both scenarios, Ansky is a slow, faint transient superposed on already ongoing accretion, which challenges the “faded AGN” interpretation invoked for some QPE hosts (Sánchez-Sáez et al., 1 Jul 2026).

6. Competing physical interpretations and unresolved constraints

No currently discussed model explains all of Ansky’s observables simultaneously. EMRI-based scenarios are attractive because they provide a natural framework for repeated interactions and can be extended to explain the rapidly evolving X-ray absorption features through shock-heated, expanding debris. They have also been used to interpret the 2024-to-2025 doubling of recurrence time and duration as a change from two-sided to one-sided stream–disk shocks, together with larger energies and longer decays. However, these scenarios encounter difficulty when confronted with the large, stable, positive $\sim 10.5$6 (Chakraborty et al., 9 Apr 2025, Hernández-García et al., 19 Sep 2025, Chakraborty et al., 18 Feb 2026).

The period-derivative analysis explicitly tested several possibilities and found each incomplete. Stable mass transfer in an EMRI can in principle increase the period, but only if the donor loses a substantial fraction of its mass on a short timescale. Pericenter kicks from asymmetric tidal mass loss require conditions close to complete disruption and are therefore difficult to reconcile with many repeated eruptions. Apparent timing changes from general relativistic precession or from light-travel delays in a hierarchical SMBH binary were found too weak or too fine-tuned, and accretion-disk instability models, while flexible in recurrence time, do not straightforwardly recover the full combination of quiescent luminosity, flare asymmetry, and stable clock-like behavior (Chakraborty et al., 18 Feb 2026).

The UV lag adds a second major constraint. Any viable model for Ansky must account for a recurrent UV counterpart, a measured delay of about one day, a moderate X-ray–UV correlation of $\sim 10.5$7, and the steadily increasing recurrence period. Star–disk collision models can reproduce the lag through diffusion in a heated blob, but they require large disk density and efficient photon thermalization. Mass-transfer and disk-instability models can reproduce the lag through light-crossing and reprocessing, but they face difficulties with the light-curve shape and the monotonic period increase (Guo et al., 3 Mar 2026).

The TDE interpretation is strengthened by the featureless HST UV spectrum, the steep continuum, the lack of broad lines, the low peak blackbody luminosity, and the exceptionally long rise and decline timescales. A specific proposal is the tidal disruption of a post-main-sequence star by a typical SMBH, which can stretch the fallback time and reduce the luminosity relative to more standard TDEs. At the same time, the host-galaxy evidence for long-lived pre-existing accretion implies that the relevant comparison is not between a TDE and a completely dormant nucleus, but between a TDE-like event and other forms of faint nuclear activity (Zhu et al., 25 Oct 2025, Sánchez-Sáez et al., 1 Jul 2026).

Taken together, the published observations place Ansky at the intersection of QPE, TDE, and low-level nuclear accretion phenomenology. A plausible implication is that the source is not a simple representative of any one class, but a system whose significance lies precisely in combining a delayed UV counterpart, evolving X-ray line formation, and a positive QPE period derivative in a nucleus with prior accretion activity. That conjunction is why Ansky has become a stringent empirical test case for models of QPE production and nuclear transient evolution (Chakraborty et al., 18 Feb 2026, Guo et al., 3 Mar 2026, Sánchez-Sáez et al., 1 Jul 2026).