What causes the runaway periodic eruptions of the nuclear transient AT2023uqm?

A second confirmed case of periodic nuclear flares

The paper reports AT2023uqm, a nuclear transient at redshift z = 0.238 that was first flagged as an outburst by ATLAS in October 2023 and later classified as an AGN-related outburst. Searching the ZTF forced-photometry light curve in September 2024 revealed three earlier, weaker outbursts, and the earliest can be traced back to June. Together with later monitoring, the source shows at least five periodic optical flares, which the authors describe as only the second confirmed case of periodicity after ASASSN-14ko.

The distinction matters because most repeated-flare candidates have shown only two flares, which is not enough to establish that the same star returns again and again. ASASSN-14ko, the best-studied comparison, has shown over 30 flares of consistent brightness. AT2023uqm differs in the opposite direction: its flares brighten over time rather than repeating at a fixed level.

Why a partial disruption can repeat on a bound orbit

In a full tidal disruption event, a star approaching a supermassive black hole too closely is torn apart, and the fallback rate of bound debris declines as t^-5/3. In a partial tidal disruption event, the star's core survives while its outer layers are stripped, and the light curve can decline more steeply. If the surviving remnant is on a bound, eccentric orbit, it can return to pericenter and be stripped again, producing periodic or recurring nuclear flares.

That is the mechanism the paper invokes for AT2023uqm. The claim is not that a single star was destroyed once, but that the same remnant repeatedly loses mass on each passage. The observed periodicity is the orbital clock; the changing flare energy is the record of how much the star has changed between passages.

The runaway energy increase is the key anomaly

The feature that separates AT2023uqm from ASASSN-14ko is a nearly exponential increase in flare energy across the sequence. The authors call this a runaway phenomenon and interpret it as a sign of the star's progressive destruction: each passage strips more material, so later flares release more energy than earlier ones.

This is an interpretation built on the flare sequence, not a direct measurement of the star's mass loss. The paper states that the behavior is consistent with repeated partial disruptions of low-mass main-sequence stars or evolved giant stars, so the runaway trend narrows the plausible stellar types without by itself choosing between them.

Multiwavelength and spectroscopic evidence for the rpTDE picture

The interpretation is reinforced by follow-up observations of the two most recent flares. The campaign covered optical and UV bands with WFST and Swift/UVOT, X-rays with Swift/XRT, Einstein Probe and XMM-Newton, and radio with e-MERLIN. WFST's deeper limiting magnitude compared with ZTF allowed a more precise characterization of the optical light curve once the source faded below about 20 mag.

Spectroscopy adds a second line of evidence. Eight spectra were obtained, including one archival spectrum from the fourth flare, with the rest taken during the fifth flare using NGPS on the Palomar 200-inch, LRIS on Keck I, the Kast double spectrograph on the Shane telescope, and BINOSPEC on the MMT. The spectra are dominated by Balmer emission lines with intermediate widths, and the paper uses the multiwavelength and spectroscopic analysis to reinforce the rpTDE interpretation. Radio observations did not detect the source, with a 5-sigma upper limit of about 100 microjansky in the first epoch.

Each flare carries a double-peaked structure

Every flare shows a similar double-peaked structure. The paper offers two possible origins: a double-peaked mass fallback rate, or two discrete collisions per orbit. These are different physical stories that happen to produce a similar light-curve shape, and the paper presents them as possibilities rather than settling the question.

The double-peaked structure also carries information about the star's size. The ratio of peak separation to orbital period is extreme, and the paper argues that if the light curve closely tracks the fallback rate at pericenter, that ratio would require a star with a radius tens of times larger than a main-sequence star. For a solar-like star disrupted by a 10^7 solar-mass black hole at impact parameter 0.6, the predicted ratio is about 2.39, far above the observed 0.038. Even the minimum value for a main-sequence star, about 2 at 1.66 solar masses, exceeds the observation.

A giant star fits the peak ratio, with caveats

Under the assumption that the light curve tracks the fallback rate, the peak-separation ratio favors an evolved giant star whose radius is tens to hundreds of times that of a main-sequence star. The paper notes that such a star has also been proposed as the disrupted star in ASASSN-14ko. Stellar spin is a further complication: a retrograde spin relative to the orbital angular momentum would narrow the energy spread of the debris, so the giant-star conclusion depends on how strongly that effect operates.

The paper does not treat the giant-star reading as settled. It states that a giant star could be distinguished from a low-mass main-sequence star by its future mass-loss evolution, which makes continued monitoring the discriminating test rather than the current data alone.

Alternative scenarios struggle with the rising flare energy

The paper also considers explanations other than a star on a bound orbit, including star-disk interaction and a massive black hole companion perturbing a pre-existing accretion disk. It disfavors them in part because they do not naturally reproduce the exponentially increasing flare energy. In the star-disk collision picture, the stripped mass would quickly saturate to a constant value over multiple collisions, and heating that expands the star would converge within less than four collisions.

A companion black hole could more naturally supply the required luminosity because of its larger interaction cross-section, but matching the large amplitude would require a companion at least as massive as an intermediate-mass black hole. Such a binary with the short orbital separation required would be short-lived before merging through gravitational-wave emission. The paper's preference for the rpTDE interpretation therefore rests on the combination of periodicity, the runaway energy trend, and the multiwavelength and spectroscopic follow-up.

Future monitoring must separate a giant from a main-sequence remnant

The paper's own framing makes the stellar type the open question, not the reality of the repeating flares. A low-mass main-sequence star and an evolved giant star can both produce the observed behavior, and the paper says the two could be distinguished by their future mass-loss evolution. That makes the next flares the decisive evidence: the runaway trend, the double-peaked structure, and the peak-separation ratio together identify AT2023uqm as the most compelling rpTDE so far, but they do not yet fix which kind of star is being cut apart.

The authors also note that the analysis is currently limited by knowledge of these events, which is why AT2023uqm is presented as a framework for modeling rather than a closed case. The priority for readers is the comparison between the two stellar scenarios and the mass-loss evolution that would separate them.

Sources used in this answer