AT 2023clx: Fast Low-Luminosity Tidal Disruption

• AT 2023clx is a tidal disruption event (TDE) in NGC 3799 marked by its proximity, extremely low luminosity, and fast light-curve evolution.
• Dense photometric, spectroscopic, and polarimetric follow-up revealed a rapid (<15 days) rise with a canonical t^(-5/3) decay, emphasizing key debris circularization processes.
• Evolving optical polarization and broad emission line profiles indicate complex outflows and shock-driven emission, challenging simple accretion-disk models.

AT 2023clx, also reported as ASASSN-23bd, is a tidal disruption event (TDE) in the nucleus of the nearby galaxy NGC 3799 at redshift $z = 0.01107$. It has been characterized as the closest optical or optical/UV TDE discovered to date, and discovery papers also identified it as the faintest or least-luminous member of the optical TDE population, although the reported peak luminosity depends on the adopted distance and host-extinction correction (Zhu et al., 2023). Because it combines proximity, low luminosity, dense photometric and spectroscopic follow-up, stringent early X-ray constraints, and later optical polarimetry, AT 2023clx has become a reference object for studies of low-luminosity, fast TDEs and of the origin of optical emission during debris circularization (Hoogendam et al., 2024).

1. Discovery, localization, and host-galaxy environment

AT 2023clx was first detected by the All-Sky Automated Survey for SuperNovae and reported on the Transient Name Server on 2023-02-26 UT; an ASAS-SN discovery analysis gives MJD 59997.2, with a $g$-band magnitude of 16.3 at discovery and a last non-detection on MJD 59988.3 (Zhu et al., 2023). Image-subtraction analysis placed the transient at the nucleus of NGC 3799 to within $0.21 \pm 0.17$ arcsec, corresponding to $\sim 49 \pm 40$ pc, which was central to its identification as a nuclear transient rather than an off-nuclear supernova (Zhu et al., 2023).

The host galaxy NGC 3799 is described as a nearby star-forming spiral galaxy and as a SB(s)b:pec system with LINER-like nuclear line ratios (Hoogendam et al., 2024). Pre-outburst optical and TESS light curves showed no significant variability over a decade, and the host showed no evidence of strong AGN activity in the last decade, although weak nuclear activity is consistent with its LINER classification (Hoogendam et al., 2024). SED-based analyses place the host on the star-forming main sequence, with reported stellar-mass estimates of $\log (M_\star/M_\odot) = 9.87 \pm 0.02$ or $M_\star \approx 6.3 \times 10^9\,M_\odot$, and one analysis found a small AGN fraction, $f_{\rm AGN}<0.2$ (Zhu et al., 2023).

Distance estimates differ modestly among studies. A discovery analysis adopted a luminosity distance of 47.8 Mpc, while another gave $50.1 \pm 3.5$ Mpc, and corrected distances up to $\sim 54.6$ Mpc were also discussed (Zhu et al., 2023). This spread propagates into the reported absolute magnitude and bolometric luminosity.

2. Light-curve evolution, rise timescale, and bolometric output

The early light curve establishes AT 2023clx as a fast, low-luminosity TDE. Fits to the rising ASAS-SN light curve indicate that it started brightening on $\mathrm{MJD}\,59988^{+1}_{-1}$, roughly 9 days before discovery, and peaked in the $g$0 band on $g$1 (Hoogendam et al., 2024). The rise was described as nearly linear in flux,

$g$2

with $g$3 (Hoogendam et al., 2024). A separate study noted that the rise was poorly sampled but consistent with a power law with $g$4 (Zhu et al., 2023). The rise time to peak was reported as $g$5 days and, after host-reddening correction, as $g$6 days; one analysis described this as the fastest-rising TDE to date (Charalampopoulos et al., 2024).

The post-peak decline was similarly notable. Optical and bolometric light curves were reported to follow the canonical $g$7 decay expected for TDE fallback (Zhu et al., 2023). Another analysis emphasized a fast 40-day bolometric decline of $g$8, placing AT 2023clx in the growing “Low Luminosity and Fast” class of TDEs (Hoogendam et al., 2024).

Reported peak photometric quantities vary across analyses:

Study	Reported peak quantity	Basis noted in study
(Zhu et al., 2023)	$g$9 mag; $0.21 \pm 0.17$0	Distance 47.8 Mpc
(Hoogendam et al., 2024)	$0.21 \pm 0.17$1	UV/optical peak luminosity
(Charalampopoulos et al., 2024)	$0.21 \pm 0.17$2 mag; $0.21 \pm 0.17$3	Host reddening corrected, $0.21 \pm 0.17$4 mag

Blackbody modeling also differs in detail among studies, but the common picture is of a blue thermal continuum with a photospheric radius of order $0.21 \pm 0.17$5 cm. One analysis fitted the optical+UV SED with the SUPERBOL package and found an almost constant blackbody temperature of $0.21 \pm 0.17$6 K, ranging from 11,000 to 13,000 K for months, with a peak radius of roughly $0.21 \pm 0.17$7 cm that decreased after peak (Zhu et al., 2023). Another reported temperatures in the range $0.21 \pm 0.17$8–20,000 K, a break in the temperature evolution during the first $0.21 \pm 0.17$9 days after peak, and a photospheric expansion velocity of $\sim 49 \pm 40$0 (Charalampopoulos et al., 2024). The bolometric luminosity in these analyses is represented by

$\sim 49 \pm 40$1

This combination of a rapid rise, low optical luminosity, and fast early decline is the basis for the classification of AT 2023clx as an extreme low-luminosity, fast TDE (Hoogendam et al., 2024).

3. Spectroscopic and multiwavelength phenomenology

Optical spectroscopy showed the characteristic TDE combination of a blue continuum with broad hydrogen and helium emission. Multiple analyses reported broad Balmer lines with widths of order $\sim 49 \pm 40$2, strong H$\sim 49 \pm 40$3, broad He II $\sim 49 \pm 40$4, and weaker He I features at 5876 and 6678 Å (Zhu et al., 2023). The spectra lacked the low-ionization metal features expected in many supernovae, arguing against a supernova interpretation (Zhu et al., 2023). H$\sim 49 \pm 40$5 profiles were asymmetric in some epochs and were attributed possibly to outflows (Zhu et al., 2023).

Several spectroscopic details are unusual even within the TDE class. One study found a flat Balmer decrement,

$\sim 49 \pm 40$6

and argued that the line emission was collisionally excited rather than produced via photoionization, in contrast to typical active galactic nuclei (Charalampopoulos et al., 2024). The same study reported a sharp, narrow emission peak at a rest wavelength of $\sim 49 \pm 40$7 Å, visible up to 10 days post-peak and interpreted as clumpy material preceding the bulk outflow, manifesting as a high-velocity component of H$\sim 49 \pm 40$8 at $\sim 49 \pm 40$9 (Charalampopoulos et al., 2024). This feature was described there as the first such case seen in TDE spectra.

Ultraviolet and near-infrared follow-up further strengthened the TDE classification. UV spectroscopy showed nitrogen emission lines such as N III] and N IV], without the strong C IV or Mg II features typical of AGN, while NIR spectroscopy lacked AGN-like broad Paschen, He I, or coronal lines (Hoogendam et al., 2024). This multiwavelength line phenomenology supported a stellar-disruption origin rather than an AGN flare (Hoogendam et al., 2024).

Early X-ray observations yielded stringent upper limits. One study reported no detection in any single or stacked Swift/XRT image, with a $\log (M_\star/M_\odot) = 9.87 \pm 0.02$0 upper limit of $\log (M_\star/M_\odot) = 9.87 \pm 0.02$1 in the 0.3–10 keV band (Zhu et al., 2023). Another gave a comparable stacked limit of $\log (M_\star/M_\odot) = 9.87 \pm 0.02$2 before MJD 60061, and also reported a late-time XMM-Newton detection on MJD 60095 of soft thermal X-rays with $\log (M_\star/M_\odot) = 9.87 \pm 0.02$3 keV and $\log (M_\star/M_\odot) = 9.87 \pm 0.02$4; more than 90% of the counts were below 2 keV (Hoogendam et al., 2024). The inferred X-ray blackbody radius of $\log (M_\star/M_\odot) = 9.87 \pm 0.02$5 cm was noted as much smaller than the Schwarzschild radius for the expected SMBH mass, a known issue for TDE X-ray spectra (Hoogendam et al., 2024).

4. Black-hole mass estimates and disruption scenarios

The central black-hole mass in NGC 3799 has been estimated with several methods, yielding a range concentrated around $\log (M_\star/M_\odot) = 9.87 \pm 0.02$6 but with non-negligible model dependence. Host-galaxy scaling in one study gave

$\log (M_\star/M_\odot) = 9.87 \pm 0.02$7

corresponding to $\log (M_\star/M_\odot) = 9.87 \pm 0.02$8 (Zhu et al., 2023). Another host-scaling estimate found

$\log (M_\star/M_\odot) = 9.87 \pm 0.02$9

while MOSFiT light-curve modeling in the same work yielded $M_\star \approx 6.3 \times 10^9\,M_\odot$0 (Hoogendam et al., 2024). A separate analysis using several methods reported $M_\star \approx 6.3 \times 10^9\,M_\odot$1 from the $M_\star \approx 6.3 \times 10^9\,M_\odot$2–$M_\star \approx 6.3 \times 10^9\,M_\odot$3 relation using an SDSS host spectrum, $M_\star \approx 6.3 \times 10^9\,M_\odot$4 from an X-shooter spectrum, $M_\star \approx 6.3 \times 10^9\,M_\odot$5 from the stellar-mass scaling, and $M_\star \approx 6.3 \times 10^9\,M_\odot$6 from MOSFiT, with an adopted average $M_\star \approx 6.3 \times 10^9\,M_\odot$7 (Charalampopoulos et al., 2024).

The host-scaling relation used in these analyses follows

$M_\star \approx 6.3 \times 10^9\,M_\odot$8

with $M_\star \approx 6.3 \times 10^9\,M_\odot$9 and $f_{\rm AGN}<0.2$0 in the formulation quoted for NGC 3799 (Zhu et al., 2023).

A later Mephisto-based study obtained systematically lower MOSFiT masses, finding

$f_{\rm AGN}<0.2$1

depending on which pre-peak data were included (Zhong et al., 2024). That same analysis identified two alternative disruption scenarios: either a full disruption of a $f_{\rm AGN}<0.2$2 star or a partial disruption of a $f_{\rm AGN}<0.2$3 star (Zhong et al., 2024). The distinction depended on inconsistencies between the ASAS-SN and ATLAS datasets during the rising phase, making the early photometry the dominant source of ambiguity (Zhong et al., 2024).

By contrast, another study argued more strongly for disruption of a very low-mass star, $f_{\rm AGN}<0.2$4, and used the rapid rise, shallow decline, and fallback-model comparison to support that interpretation (Charalampopoulos et al., 2024). In that framework, the SMBH mass of order $f_{\rm AGN}<0.2$5 ruled out an intermediate-mass black hole as the explanation for the fast rise (Charalampopoulos et al., 2024).

These differing inferences are not purely numerical disagreements; they reflect the sensitivity of TDE parameter recovery to extinction corrections, pre-peak sampling, and the assumed radiative channel.

5. Optical emission mechanism and polarimetric constraints

The central physical question for AT 2023clx is the origin of its optical radiation and the timescale of disk formation. One interpretation emphasizes efficient circularization and prompt accretion-disk formation in the disruption of a very low-mass star, viewed through a low-density photosphere and accompanied by an outflow launched in the line of sight (Charalampopoulos et al., 2024). Another interpretation, based on composite multi-band light-curve fitting and the early lack of soft X-rays, proposes that the observed optical radiation is powered by stream-stream collision rather than prompt accretion onto a compact disk (Zhong et al., 2024).

In the stream-collision picture, the self-intersection radius is at $f_{\rm AGN}<0.2$6–$f_{\rm AGN}<0.2$7 cm, the fitted radiative efficiencies are low, $f_{\rm AGN}<0.2$8–$f_{\rm AGN}<0.2$9, and the lack of early X-rays follows from delayed circularization (Zhong et al., 2024). Using the orbital period of the most bound debris, $50.1 \pm 3.5$0–28 days, and a collision efficiency $50.1 \pm 3.5$1, that study estimated circularization timescales of $50.1 \pm 3.5$2 days for the low-mass full-disruption solutions and $50.1 \pm 3.5$3 days for the near-solar-mass partial-disruption solutions, and therefore speculated that soft X-rays may emerge 100–600 days after the optical peak (Zhong et al., 2024). This is an explicitly model-dependent prediction.

Optical polarimetry has provided the strongest direct geometric constraint. Nordic Optical Telescope observations covered five epochs from near optical peak, $50.1 \pm 3.5$4 days after maximum, to about 35 days post-peak, with later upper limits (Koljonen et al., 12 Aug 2025). The linear Stokes parameters were derived as

$50.1 \pm 3.5$5

with polarization degree

$50.1 \pm 3.5$6

and polarization angle

$50.1 \pm 3.5$7

The polarimetric evolution was highly structured. The earliest observation, 6 days after peak, showed low or undetectable polarization in $50.1 \pm 3.5$8 band, $50.1 \pm 3.5$9, while $\sim 54.6$0- and $\sim 54.6$1-band values were at $\sim 54.6$2 and $\sim 54.6$3 (Koljonen et al., 12 Aug 2025). The $\sim 54.6$4- and $\sim 54.6$5-band polarization degrees then increased nearly linearly until $\sim 54.6$6 days after peak, with a linear growth rate of $\sim 54.6$7 per day, reaching $\sim 54.6$8 in $\sim 54.6$9 band and $\mathrm{MJD}\,59988^{+1}_{-1}$0 in $\mathrm{MJD}\,59988^{+1}_{-1}$1 band on day 35.7 (Koljonen et al., 12 Aug 2025). After $\mathrm{MJD}\,59988^{+1}_{-1}$2 days the polarization degree fell below 1.5% in $\mathrm{MJD}\,59988^{+1}_{-1}$3 band at the $\mathrm{MJD}\,59988^{+1}_{-1}$4 level (Koljonen et al., 12 Aug 2025). The polarization angle was around $\mathrm{MJD}\,59988^{+1}_{-1}$5 in $\mathrm{MJD}\,59988^{+1}_{-1}$6 at 6 days post-peak, dropped by about $\mathrm{MJD}\,59988^{+1}_{-1}$7–$\mathrm{MJD}\,59988^{+1}_{-1}$8 over the next two epochs to $\mathrm{MJD}\,59988^{+1}_{-1}$9–$g$00 at days 20–36, and then remained relatively stable (Koljonen et al., 12 Aug 2025). The wavelength dependence also evolved from red-dominated polarization to blue-dominated polarization (Koljonen et al., 12 Aug 2025).

This variability is difficult to reconcile with simple reprocessing models, which were described as predicting minimal, slow polarization variability and generally lower polarization degree, $g$01 even for favorable geometries (Koljonen et al., 12 Aug 2025). A collisionally-induced outflow model can allow high and time-variable polarization degrees and a rise and fall in polarization, but the study noted that this scenario does not currently address polarization-angle evolution, which is the most distinctive observable in the data (Koljonen et al., 12 Aug 2025). The close resemblance to the polarization evolution of AT 2020mot was therefore taken as strong evidence that tidal stream shocks dominate the optical outburst during accretion-disk formation (Koljonen et al., 12 Aug 2025). This suggests that AT 2023clx probes the circularization stage directly rather than only a later reprocessing layer.

6. Position within the TDE population and outstanding issues

AT 2023clx occupies an important region of TDE parameter space. Discovery studies described it as the faintest or least-luminous optical TDE yet found and as the nearest event of its class, while later work emphasized its membership in the “Low Luminosity and Fast” population (Zhu et al., 2023). Its host environment is also atypical relative to the overrepresentation of post-starburst hosts in historical TDE samples: NGC 3799 is a star-forming main-sequence spiral with LINER-like nuclear emission (Zhu et al., 2023).

Population-level implications have already been drawn. One luminosity-function analysis concluded that adding AT 2023clx doubled the volumetric rate at $g$02 relative to previous ZTF-based estimates and suggested that faint TDEs may constitute up to 74% of all TDEs in the observed $g$03-band peak-luminosity range $g$04 (Zhu et al., 2023). This suggests that flux-limited surveys have likely missed a substantial nearby population of faint TDEs.

Several issues remain unsettled. The adopted host-extinction correction changes the absolute magnitude and bolometric luminosity significantly (Charalampopoulos et al., 2024). Black-hole mass estimates range from $g$05 to $g$06 depending on the method and dataset (Zhong et al., 2024). The disrupted star may have been a very low-mass star, or the light curve may permit a partial disruption of a nearly solar-mass star (Zhong et al., 2024). Most importantly, the prompt-disk and delayed-circularization pictures are not equivalent: they imply different locations of the dominant optical emitter and different expectations for late-time X-ray emergence (Charalampopoulos et al., 2024). High-cadence, multi-band early photometry, continued X-ray monitoring, and additional time-resolved polarimetry are therefore central to resolving the physical interpretation.

In this sense, AT 2023clx is not only an unusually nearby and faint TDE; it is also a stringent test case for how optical TDE emission is produced during the transition from stellar disruption to disk formation.