11email: ashley.chrimes@esa.int 22institutetext: Department of Astrophysics/IMAPP, Radboud University, PO Box 9010, 6500 GL Nijmegen, The Netherlands 33institutetext: Department of Physics, University of Warwick, Gibbet Hill Road, CV4 7AL Coventry, United Kingdom 44institutetext: Inter-University Institute for Data Intensive Astronomy, Department of Astronomy, University of Cape Town, Private Bag X3, Rondebosch 7701, South Africa 55institutetext: South African Astronomical Observatory, P.O. Box 9, 7935 Observatory, South Africa 66institutetext: Oxford Astrophysics, Denys Wilkinson Building, Keble Road, Oxford, OX1 3RH, United Kingdom
Multi-wavelength observations of the Luminous Fast Blue Optical Transient AT 2023fhn
Abstract
Context. Luminous Fast Blue Optical Transients (LFBOTs) are a class of extragalactic transients notable for their rapid rise and fade times, blue colour and accompanying luminous X-ray and radio emission. Only a handful have been studied in detail since the prototypical example AT 2018cow. Their origins are currently unknown, but ongoing observations of previous and new events are placing ever stronger constraints on their progenitors.
Aims. We aim to put further constraints on the LFBOT AT 2023fhn, and LFBOTs as a class, using information from the multi-wavelength transient light-curve, its host galaxy and local environment.
Methods. Our primary results are obtained by fitting galaxy models to the spectral energy distribution of AT 2023fhn’s host and local environment, and by modelling the radio light-curve of AT 2023fhn as due to synchrotron self-absorbed emission from an expanding blast-wave in the circumstellar medium.
Results. We find that the neither the host galaxy nor circumstellar environment of AT 2023fhn are unusual compared with previous LFBOTs, but that AT 2023fhn has a much lower X-ray to ultraviolet luminosity ratio than previous events.
Conclusions. We argue that the variety in ultraviolet-optical to X-ray luminosity ratios among LFBOTs is likely due to viewing angle differences, and that the diffuse, yet young local environment of AT 2023fhn - combined with a similar circumstellar medium to previous events - favours a progenitor system containing a massive star with strong winds. Plausible progenitor models in this interpretation therefore include black hole/Wolf-Rayet mergers or failed supernovae.
Key Words.:
Supernovae: individual: AT 2023fhn – Supernovae: general – Stars: black holes – Black hole physics – Stars: winds, outflows – Stars: circumstellar matter1 Introduction
Luminous Fast Blue Optical Transients (LFBOTs) are a rare class of rapidly-evolving, hot, multi-wavelength extragalactic transients. The prototypical example, AT 2018cow (‘the Cow’, Prentice et al., 2018), is the nearest and best-studied event of this class so far. Its characteristic early-time features include a peak optical absolute magnitude of –20 with a rapid rise and decay timescale of 5 days, constraining the 56Ni mass in the ejecta to M⊙, and ruling out standard core-collapse or thermonuclear supernova models (Perley et al., 2019; Margutti et al., 2019). The optical-ultraviolet (UV) emission is well fit by a hot black-body and power-law component, where the black-body temperature was K initially, falling to K over two weeks (Prentice et al., 2018; Perley et al., 2019). The optical spectra were largely featureless, with broad hydrogen absorption features (indicative of a high outflow velocity) appearing and disappearing between 2–8 days, and narrow He lines appearing after 20 days. At other wavelengths, AT 2018cow was X-ray and radio bright (Rivera Sandoval et al., 2018; Margutti et al., 2019; Ho et al., 2019; Nayana & Chandra, 2021). The X-ray emission was well in excess of power-law extrapolations from the radio (e.g. Ho et al., 2019) and was also highly variable after a break in the light-curve, which declined as after 20 days (Migliori et al., 2024). The broadband X-ray spectrum, and X-ray variability, cannot be explained by an external shock origin, and the synchrotron self-absorbed radio emission - consisting of a slow rise and rapid decay - was instead smoothly evolving, indicating a distinct physical origin from the highly variable X-rays. An interpretation is that a central engine powers the X-ray emission, while an expanding blast wave produces the radio emission (e.g. Ho et al., 2019; Margutti et al., 2019). The slow radio variability timescale set the size of the emission region at 5–6 days at cm, while the X-ray variability gave a length scale five times smaller (Ho et al., 2019). Therefore, the X-rays appear to originate from a central engine or internal shock, while the radio emission is generated externally. A claim of quasi-periodic oscillations in the X-rays can interpreted as evidence for a central engine (Pasham et al., 2021), while a separate claim of 250s quasi-periodicity instead implies an intermediate mass (103-105 M⊙) black hole (Zhang et al., 2022). Synchrotron modelling of the sub-millimetre and radio data revealed a mildly-relativistic expansion velocity (0.1) into a wind-like extended circumstellar medium (CSM) with a high density of 105 cm-3 (Margutti et al., 2019; Ho et al., 2019). For Wolf-Rayet-like wind speeds of 1000 km s-1, this implies a mass-loss rate - yr-1 (Margutti et al., 2019).
Since AT 2018cow, several more LFBOTs have been discovered. Confirmed events include AT 2018lug/ZTF 18abvkwla (‘the Koala’, Ho et al., 2020), CSS161010 (Coppejans et al., 2020), AT 2020xnd/ZTF 20 acigmel (‘the Camel’, Perley et al., 2021; Bright et al., 2022; Ho et al., 2022), AT 2020mrf (Yao et al., 2022), AT 2022tsd (‘the Tasmanian Devil’, Matthews et al., 2023) and AT 2023fhn (‘the Finch’, Chrimes et al., 2024). Despite variety (e.g. in peak luminosity), they share the same key features of hot, largely featureless spectra at early times, optical luminosities rivalling gamma-ray burst (GRB) afterglows and superluminous supernovae, plus bright X-ray and radio emission. They are estimated to occur at 0.1% of the local core-collapse supernova rate (Ho et al., 2023d).
Recent developments have provided further insight into the origin of LFBOTs. Polarimetry of AT 2018cow demonstrated the emission region to be highly aspherical, indicative of an accretion disc (Maund et al., 2023). Unexpectedly, AT 2018cow was found to be UV (Sun et al., 2022; Chen et al., 2023; Sun et al., 2023; Inkenhaag et al., 2023) and X-ray (Migliori et al., 2024) bright at late times, several years post-explosion. This emission has been interpreted as from a black hole accretion disc. Estimates for the black hole mass range from 10–100 M⊙ (super-Eddington accretion) to 103–104 M⊙ (sub-Eddington, from X-ray observations, Migliori et al., 2024) and 1000 M⊙ (UV observations, Inkenhaag et al., 2023). Magnetar central engine models struggle to produce both the early and late UV emission (Chen et al., 2023). Further evidence for a black hole accretion scenario comes from minute-long optical flares, up to several months post-explosion, from AT 2022tsd (Ho et al., 2023c). An interpretation is that the central engine is undergoing highly variable, short-lived bursts of accretion.
Several models have been put forward to explain LFBOTs. Tidal disruptions of compact, hydrogen-poor stars (such as white dwarfs) around intermediate mass black holes (IMBHs) can plausibly explain the optical rise and fall timescale, spectral features and X-ray variability timescale (Perley et al., 2019; Kuin et al., 2019). However, the dense CSM inferred from radio observations is hard to explain in such a scenario (e.g. Margutti et al., 2019). Other possibilities include failed supernovae, in which a black hole is formed and the emission is powered by accretion onto the natal black hole rather than radioactive decay in the ejecta (Perley et al., 2019; Quataert et al., 2019), choked jets (e.g. Gottlieb et al., 2022; Soker, 2022), highly aspherical supernovae (‘ellipsars’, DuPont et al., 2022), and the mergers of compact objects and/or massive stars (Lyutikov & Toonen, 2019; Uno & Maeda, 2020; Schrøder et al., 2020), such as black holes and Wolf-Rayet stars (Metzger, 2022). A dense outflow from the progenitor may result in dust echoes (Metzger & Perley, 2023). CSM shock interaction models have also been put forward (e.g. Fox & Smith, 2019; Xiang et al., 2021; Pellegrino et al., 2022; Khatami & Kasen, 2023), but the X-ray variability, broadband spectral evolution, late-time UV/X-ray emission from AT 2018cow and giant optical flares from AT 2022tsd all indicate the presence of a central engine.
In this paper, we present multi-wavelength radio, optical, UV, and X-ray observations of the LFBOT AT 2023fhn up to 200 days post-explosion. We place AT 2023fhn in the context of other LFBOTs so far, in terms of its host galaxy, optical/UV/X-ray light-curve, and radio emission, with the event energetics and blast wave properties inferred from synchrotron modelling of the radio observations. Throughout, we use a flat CDM cosmology with =0.3 and =70 kms-1Mpc-1. All magnitudes are reported in the AB system (Oke & Gunn, 1982).
2 Observations and data reduction
2.1 X-ray
We obtained four epochs of Chandra X-ray Observatory (CXO) ACIS-S observations of AT 2023fhn up to 200 days post-explosion. The epochs consist of 1, 2, 6 and 14 observations, respectively (full details are provided in Table 1). The data are reduced, and transient fluxes measured, with standard CIAO (v4.13, caldb v4.9.3, Fruscione et al., 2006) procedures. The images are reprocessed and filtered to the energy range 0.5-7.0 keV. wavdetect is used to find point sources, and srcflux used to measure the flux (or upper limits) at the location of AT 2023fhn. We merged the datasets in each of the four epochs (with mergeobs) to increase the signal-to-noise ratio. The mean (mid-point, exposure-time weighted) observation times of these epochs are 15.0, 28.9, 64.5 and 210.9 days (since JD–2460045, or 12:00 UT on 10-Apr-2023). The total exposure times per epoch are 30, 60, 83 and 193 ks respectively. Finally, the fluxes are de-absorbed by assuming a photon index (e.g. Rivera Sandoval et al., 2018), and a Galactic neutral hydrogen column density of cm-2 (Dickey & Lockman, 1990).
ObsID | Start date | texp | Data mode | FX |
---|---|---|---|---|
JD-2460045 | ks | erg s-1 cm-2 | ||
26624 | 14.78957 | 29.68 | FAINT | (7.6) |
26625 | 27.98310 | 29.68 | FAINT | (4.5) |
27833 | 29.47145 | 29.67 | FAINT | |
26626 | 61.80356 | 16.88 | VFAINT | |
27895 | 62.33516 | 10.94 | VFAINT | |
27835 | 65.12251 | 13.89 | FAINT | |
27905 | 65.45429 | 13.89 | FAINT | |
27906 | 65.79704 | 13.89 | FAINT | |
27907 | 66.13969 | 13.89 | FAINT | |
26627 | 198.66317 | 10.74 | VFAINT | |
28997 | 198.96634 | 10.74 | VFAINT | |
28998 | 199.27127 | 11.12 | VFAINT | |
27837 | 205.89744 | 13.4 | VFAINT | |
29031 | 206.23509 | 13.3 | VFAINT | |
29032 | 206.57170 | 13.5 | VFAINT | |
29034 | 207.96147 | 10.93 | VFAINT | |
29033 | 208.26639 | 14.39 | VFAINT | |
27838 | 215.80840 | 16.85 | VFAINT | |
29054 | 216.19957 | 17.84 | VFAINT | |
29056 | 216.60163 | 14.39 | VFAINT | |
28991 | 218.76822 | 9.94 | VFAINT | |
28999 | 219.12318 | 18.69 | VFAINT | |
29055 | 219.54352 | 16.85 | VFAINT |
2.2 UV-optical
A second epoch of HST imaging was obtained on 23/24 October 2023 (the first was on 17 May 2023, Chrimes et al., 2024), using the WCF3 instrument and six filters (). Full details are given in Table 2. The data are reduced with drizzlepac (Hoffmann et al., 2021), re-drizzling the charge-transfer-efficiency-corrected flc input images with North oriented up and a final pixel scale of 0.025 arcsec pixel-1 (pixfrac=0.8). Image stamps around the location of AT 2023fhn in epochs 1 and 2 are shown in Figure 1. Visible in the bottom left is the presumed satellite of the larger spiral to the south (see Figure 3). Both galaxies lie at a common redshift of 0.24 (Ho et al., 2023b; Chrimes et al., 2024).
Filter | Start date | texp |
---|---|---|
JD–2460045 | s | |
F555W | 36.87666 | 1092 |
F814W | 36.89272 | 1092 |
F555W | 196.42824 | 990 |
F814W | 196.44313 | 1092 |
F225W | 196.49431 | 1068 |
F336W | 196.51027 | 1068 |
F845M | 196.56034 | 990 |
F763M | 196.57515 | 1068 |
2.3 Radio
We obtained radio observations with the Karl G. Jansky Very Large Array (VLA) between 22 Apr 2023 and 16 December 2024 (programme SC240143, PI: Chrimes). Details of the observations are listed in Table 3. The observations were taken in standard phase-referencing mode using 3C286 as a flux density and bandpass calibrator, with ICRF J101447.0+230116, FIRST J101644.3+203747, FIRST J101353.4+244916 and ICRF J095649.8+251516 as complex gain calibrators. The observations were calibrated using the VLA Calibration Pipeline versions 2023.1.0.124 and 2022.2.0.64 in CASA 6.5.4 and 6.4.1 respectively, with additional manual flagging. The images were created using the tclean task in CASA with Briggs weighting with a robust parameter of 1. In the observations where the source was not detected we quote the upper limit on the flux density as three times the local RMS. The one exception to this is during the last epoch (see Table 3) where the synthesized beam (resolution element) was large and included other sources. In this case we quoted the upper limit as the flux density at the source location. For the observations where we detected the target, we fitted the flux density using the imfit task within CASA and constrained the fit to the synthesized beam.
The observations up to 12 days post JD-2460045 are already published (Chrimes et al., 2024) and all produced non-detections. In the 87–95 day and day epochs we have sufficient data points for fitting a synchrotron self-absorbed spectrum. The KU band (15 GHz) data point at 138 days has sufficient signal-to-noise to split into 3 (centred on 13, 15 and 17 GHz), as listed in Table 3, increasing the points at 138 days to 7 (with 6 detections). We fit a self-absorbed synchrotron model to the 87-95 and 138 day epochs in Section 4.4.
Start date | Freq. | Bandwidth | texp | Flux Density |
---|---|---|---|---|
JD–2460045 | GHz | GHz | min. | |
11.80740 | 1.52 | 1.024 | 38 | 130 |
11.78309 | 3.00 | 2.048 | 32 | 35 |
11.76507 | 6.00 | 4.096 | 23 | 18 |
11.74688 | 10.00 | 4.096 | 23 | 18 |
11.72090 | 15.08 | 6.144 | 35 | 11 |
11.69229 | 22.00 | 8.192 | 35 | 17 |
11.66552 | 33.00 | 8.192 | 33 | 25 |
87.59185 | 1.52 | 1.024 | 39 | 45 |
87.56657 | 3.00 | 2.048 | 33 | 1108 |
87.54257 | 6.00 | 4.096 | 32 | 1285 |
95.58690 | 10.00 | 4.096 | 31 | 1057 |
95.56438 | 15.00 | 6.144 | 30 | 717 |
95.52738 | 22.00 | 8.192 | 47 | 6010 |
137.17440 | 1.52 | 1.024 | 39 | 58 |
137.14178 | 3.00 | 2.048 | 44 | 11010 |
137.10567 | 6.00 | 4.096 | 50 | 2215 |
138.16972 | 10.00 | 4.096 | 42 | 1976 |
138.13664 | 13.00 | 2.050 | 42 | 18010 |
138.13664 | 15.00 | 2.050 | 42 | 16010 |
138.13664 | 16.96 | 2.050 | 42 | 14020 |
249.95139 | 6.00 | 4.096 | 35 | 311a |
249.90997 | 10.00 | 4.096 | 57 | 181a |
249.86861 | 15.08 | 6.144 | 57 | 125a |
3 Environmental analysis
3.1 Local environment
The second epoch of HST imaging presented in this paper allows us to examine the environment directly underlying the transient after it has faded. As noted by Chrimes et al. (2024), there is diffuse emission in the vicinity of the transient. To characterise this faint underlying population, we place 0.2 arcsec (and 0.4 arcsec) apertures at the location of AT 2023fhn in all six epoch 2 images. The images are aligned with - shifts using 5 common point sources in every image, with respect to the location of AT 2023fhn in the epoch 1 F555W image. The rms of these relative astrometric alignments is 5–10 mas, better than the absolute astrometry of the images (which have been aligned with the Gaia DR3 reference frame), and much smaller than the aperture size. We perform photometry with photutils, estimating the background with either the median image background (with medianbackground) or an annulus (1.5 to 4 times the aperture radius, with pixels values clipped at 3). The appropriate encircled energy corrections for each filter and aperture are applied. Magnitudes are then calculated using the photplam and photflam header keywords111https://hst-docs.stsci.edu/wfc3dhb/chapter-9-wfc3-data-analysis/9-1-photometry, and are listed in Table 4. The only detections are in F555W and F814W. To investigate the nature of these detections, we place eight 0.4 arcsec apertures at equal spacing around the location of AT 2023fhn in a circle of radius 20 pixels (0.5 arcsec). With the F555W filter and median background subtraction, we have significant detections in 5/8 apertures, with a mean magnitude of 25.90.6 in these apertures - consistent with the measurement at the precise location of AT 2023fhn. This demonstrates that the emission in this area is from an extended, diffuse background, rather than any significant contribution by residual light from AT 2023fhn. This can also be seen in Table 4, where the magnitudes calculated with annulus background subtraction are fainter, since the local background is elevated. Larger apertures also give brighter magnitudes, despite encircled energy correction (unlike point sources in the field). We similarly disfavour any significant contribution from a compact cluster at this specific location, which would appear as a point source in the image given the physical scale at this redshift of 100 pc pixel-1. However, the presence of a globular cluster (which would favour an IMBH interpretation, e.g., Lützgendorf et al., 2013) cannot be ruled out, as even the brightest globular clusters would be far below detection limits at this distance and limiting magnitude (Chrimes et al., 2024). Shifting the circle of apertures 5 arcsec to the north, well away from the galaxies, we find non-detections in all eight apertures with a 3 upper limit of 26.7. We therefore conclude that there is extended, diffuse emission from an underlying stellar population at the location of AT 2023fhn.
We now estimate the age and dust extinction of this underlying population. First, we correct for the (low) Galactic extinction of =0.0254 (Schlafly & Finkbeiner, 2011)222https://irsa.ipac.caltech.edu/applications/DUST/ using the filter effective wavelengths (Rodrigo et al., 2012; Rodrigo & Solano, 2020) and the Python extinction package (Barbary, 2016) with a Fitzpatrick (1999) extinction law and . To estimate the age and local (intrinsic) extinction, we fit the Galactic-extinction corrected and photometry to BPASS (Binary Population and Spectral Synthesis v2.1, Eldridge et al., 2017; Stanway & Eldridge, 2018) single-age spectral templates. These are constructed by assuming that a stellar population of 106M⊙ is formed instantaneously, and left to evolve with no further star formation. We use these simple stellar populations since the limited data available to model solely the local environment of AT 2023fhn precludes a more complex procedure, including, for example, the star-formation history (however, see the next section). A fixed metallicity of half-Solar is adopted ( by mass fraction). We therefore simply fit for the age of the population, the luminosity (i.e. mass) of the stellar population is then allowed to freely vary to minimise . Four data points are used ( and ) where the upper limits are treated as data points with zero flux and an uncertainty equal to the flux of the 1 upper limit. We therefore have 2 fit parameters and 4 data points for 2 degrees of freedom. Fitting is performed by multiplying the (de-redshifted) filter response curves (Rodrigo et al., 2012; Rodrigo & Solano, 2020) with the BPASS spectra to extract fluxes and hence magnitudes from the spectra. These are compared with the absolute magnitudes in each filter, after correction for a range of intrinsic extinction values from =0.0 to 1.0. The intrinsic extinction correction uses the rest-frame effective wavelength of each filter. The and filters are not used in this fit since the upper limits are shallower than the and detections, and so provide no additional constraints.
The results are shown in Figure 2. The top panel shows the best-fit single-age BPASS spectrum. The lower panel shows log across the parameter space (indicated by the shading). Each pixel represents a unique combination of and a BPASS simple stellar population at a given age. The 68% and 90% confidence intervals are indicated by white contours (where the intervals are from Avni, 1976).
We also measure the local surface brightness in epoch 2 (in a 0.5 arcsec radius around AT 2023fhn’s position), giving 25.1 mag arcsec-2 in and 24.65 mag arcsec-2 in . This compares well with the 25.2 mag arcsec-2 and 24.6 mag arcsec-2 values from the transient-subtracted images in Epoch 1 (see Chrimes et al., 2024). The surface brightness is 25.76 mag arcsec-2, which after Galactic extinction correction is 25.27 mag arcsec-2. The rest-frame central wavelength of is 2700Å. This allows for a better comparison with the UV () surface brightness distribution for supernova environments, as reported by Kelly & Kirshner (2012) than made by Chrimes et al. (2024) with . The Galactic extinction-corrected surface brightness is in the faintest 10% for local supernova values; this is therefore faint but not unprecedented. We note that supernovae type IIb are the most likely supernova sub-class to explode in young, but low surface brightness environments, and are also found at the highest host-normalised offsets on average (Kelly & Kirshner, 2012).
Filter | Method | Bkg. | Aper. | m | m |
---|---|---|---|---|---|
F225W | photutils | Median | 0.2′′ | ¿26.1 | - |
F225W | photutils | Annulus | 0.2′′ | ¿26.1 | - |
F225W | photutils | Median | 0.4′′ | ¿25.4 | - |
F225W | photutils | Annulus | 0.4′′ | ¿25.5 | - |
F336W | photutils | Median | 0.2′′ | ¿26.6 | - |
F336W | photutils | Annulus | 0.2′′ | ¿26.6 | - |
F336W | photutils | Median | 0.4′′ | ¿25.9 | - |
F336W | photutils | Annulus | 0.4′′ | ¿25.9 | - |
F555W | photutils | Median | 0.2′′ | 26.9 | 0.2 |
F555W | photutils | Annulus | 0.2′′ | 27.1 | 0.3 |
F555W | photutils | Median | 0.4′′ | 25.8 | 0.2 |
F555W | photutils | Annulus | 0.4′′ | 25.6 | 0.1 |
F763M | photutils | Median | 0.2′′ | ¿26.0 | - |
F763M | photutils | Annulus | 0.2′′ | ¿26.0 | - |
F763M | photutils | Median | 0.4′′ | ¿25.3 | - |
F763M | photutils | Annulus | 0.4′′ | 25.0 | 0.3 |
F814W | photutils | Median | 0.2′′ | 26.4 | 0.2 |
F814W | photutils | Annulus | 0.2′′ | 26.5 | 0.3 |
F814W | photutils | Median | 0.4′′ | 25.3 | 0.2 |
F814W | photutils | Annulus | 0.4′′ | 25.2 | 0.2 |
F845M | photutils | Median | 0.2′′ | ¿25.6 | - |
F845M | photutils | Annulus | 0.2′′ | ¿25.6 | - |
F845M | photutils | Median | 0.4′′ | ¿24.9 | - |
F845M | photutils | Annulus | 0.4′′ | ¿24.9 | - |
3.2 Global host properties
We next consider how the overall host galaxy properties compare with the local environment of AT 2023fhn, and how they compare with the hosts of other LFBOTs. To do this, we perform spectral energy distribution (SED) fitting of the integrated light of the host. By host, we refer to the spiral and satellite galaxy together, since their proximity likely results in interactions (e.g. tidal) and therefore the two galaxies can be considered as one interacting system. Furthermore, the two galaxies are not spatially resolved in ground-based imaging (e.g. PanSTARRS), which we used to add photometric points to the SED.
We attempt to collect as close to 100% of the galaxy light from HST photometry as possible. We measure the Petrosian radius (Petrosian, 1976) of the spiral galaxy with the statmorph package (Rodriguez-Gomez et al., 2019, ) and adopt 1.5 as a radius that encloses 100% of the flux (e.g. Conselice, 2003). We account for the projected ellipticity and orientation of the galaxy using the ellip and theta outputs. A pixel mask is produced using these parameters as measured from the image, and applied to the other HST images, as shown in Figure 3. The flux within the mask is summed, and background subtraction (as for the local environment measurements above) uses the sigma-clipped median background, scaled for the number of pixels in the mask. Repeating the procedure for the satellite galaxy produces a 1.5 pixel mask that lies entirely within the spiral’s mask. We therefore use spiral pixel mask alone as it captures 100% of the flux from both galaxies.
To supplement the HST data we add host photometry from archival catalogues. For additional optical points we use PanSTARRS data release 2 (Chambers et al., 2016). We use the catalogued Kron magnitudes (Kron, 1980, in , , , and ), which capture 90% of the light of extended sources, and increase the fluxes by a further 10% to approximate the 100% flux value 333https://outerspace.stsci.edu/display/PANSTARRS/PS1+Kron+photometry+of+extended+sources. The Kron radii for the spiral (4.41, 4.62, 4.36, 3.33 and 2.89 arcsec in respectively) extend past the position of the satellite in , so the system can be considered blended. Effective wavelengths for these filters are from Tonry et al. (2012). We also add far-UV and near-UV photometry from GALEX (Martin et al., 2003), plus W1, W2 and W3 detections from WISE. The spiral and satellite cannot be separated at the spatial resolution of these surveys, and neither galaxy is detected in 2MASS. The full list of photometry used to performed SED fitting is provided in Table 5.
To perform SED fitting we use prospector (Leja et al., 2017; Johnson et al., 2021), which makes use of FSPS (Flexible Stellar Population Synthesis Conroy et al., 2009; Conroy & Gunn, 2010) and Python-FSPS (Johnson et al., 2023). For the Markov Chain Monte Carlo (MCMC) implementation we use emcee (Foreman-Mackey et al., 2013). We again use BPASS (Binary Population and Spectral Synthesis v2.1, Eldridge et al., 2017; Stanway & Eldridge, 2018) for the spectral models. Before being passed to prospector, the input photometry is corrected for Galactic extinction (as described in Section 3.1). We fit four parameters: the stellar mass , intrinsic extinction , population age and the timescale for an exponentially declining star-formation history . The redshift is fixed at , and the luminosity distance at Mpc.
We run the MCMC with 128 walkers and 512 iterations; the full list of MCMC set-up parameters and joint posterior distributions (in the form of a corner plot) are provided in Appendix A. The maximum a posterior (MAP) spectrum is shown in Figure 4, with the associated properties from the posterior distribution listed in Table 6. Thus far, the metallicity has been fixed at half-Solar, based on the approximate mass of M⊙ and the mass-metallicity relation (Tremonti et al., 2004; Gallazzi et al., 2005). A similar table containing the results when metallicity is allowed to vary is also provided in Appendix A. In this case, the mass and SFR are similar, such that fixing at a more realistic value does not change our results in a qualitative sense. In the delayed- model, the current star-formation rate (SFR) is proportional to . The absolute value is obtained by normalisation with respect to the mass formed, yielding a SFR of 4 M⊙ yr-1. The galaxy pair is therefore dominated by a fairly typical star-forming spiral, but is perhaps notable for the likely presence of tidal interactions between the spiral and its satellite. In Figure 5 we plot its mass versus SFR, comparing with the host galaxies of previous LFBOTs. The galaxy has a high SFR and mass for LFBOT hosts, lying slightly above average in terms of specific star formation rate (sSFR), but well below the sSFR of the host of ZTF 18abvkwla.
Filter | Source | m | err | [Å] | A() |
---|---|---|---|---|---|
FUV | GALEX | 20.93 | 0.31 | 1548.85 | 0.20 |
NUV | GALEX | 20.74 | 0.25 | 2303.37 | 0.22 |
F225W | HST | 20.60 | 0.05 | 2358.70 | 0.20 |
F336W | HST | 20.40 | 0.03 | 3359.11 | 0.13 |
PS | 19.70 | 0.01 | 4810.00 | 0.09 | |
F555W | HST | 19.34 | 0.01 | 5235.33 | 0.08 |
PS | 19.17 | 0.01 | 6170.00 | 0.07 | |
PS | 18.93 | 0.01 | 7520.00 | 0.05 | |
F763M | HST | 18.93 | 0.01 | 7602.85 | 0.05 |
F814W | HST | 18.84 | 0.01 | 7954.84 | 0.04 |
F845M | HST | 18.68 | 0.01 | 8430.20 | 0.04 |
PS | 19.02 | 0.02 | 8660.00 | 0.04 | |
PS | 19.00 | 0.04 | 9620.00 | 0.03 | |
W1 | WISE | 18.91 | 0.07 | 33526.00 | 0.00 |
W2 | WISE | 18.82 | 0.13 | 46028.00 | 0.00 |
W3 | WISE | 16.92 | 0.38 | 115608.00 | 0.00 |
Host property | Value |
---|---|
/ M⊙ | |
SFR / M⊙ yr-1 | 4.0 |
0.15 | |
/Gyr | 0.95 |
/Gyr | 0.26 |
4 Transient emission
4.1 UV-optical
We now compare the UV-optical constraints on AT 2023fhn’s light-curve with previous LFBOTs. All times used in Section 4 are in the rest-frames of the LFBOTs. Comparison data are corrected for Galactic extinction of =0.08 (AT 2018cow, Prentice et al., 2018) and =0.07 (ZTF 20acigmel, Perley et al., 2021), their UV light-curves (in absolute magnitude) are compared with AT 2023fhn in Figure 6. We fit the light-curve of AT 2018cow in 2 phases, early ( d) and late-time, with a fit of the form . For the fit to AT 2018cow, we assume that the late-time UV is dominated by residual transient emission (Sun et al., 2022, 2023; Chen et al., 2023; Inkenhaag et al., 2023). We shift the AT 2018cow best-fit up in absolute magnitude such that it lies between the early-time ATLAS -band and FORS2 -band AT 2023fhn detections (Ho et al., 2023b). The extrapolated curve passes below the late-time HST F225W and F336W upper limits reported in this work. Another LFBOT with good UV photometric coverage is ZTF 20acigmel, but here we consider only the early, pre-break phase due to a lack of late-time constraints. ZTF 20acigmel starts brighter than AT 2018cow and fades faster, whereas AT 2023fhn is the most luminous LFBOT yet at UV-optical wavelengths. A final addition to Figure 6 are bands of constant UV absolute magnitude, corresponding to late-time emission from black holes of different masses in the tidal disruption event model of Mummery et al. (2024). This model yielded a black hole mass of 103 M⊙ for AT 2018cow. Assuming similar evolution, the HST F336W point source upper limit for AT 2023fhn tentatively constrains the accreting black hole mass in a TDE interpretation to M⊙.
4.2 X-ray
Figure 7 shows our X-ray observations of AT 2023fhn, and the X-ray light-curves of other LFBOTs. The AT 2018cow broken power-law and late-time plateau fit of Migliori et al. (2024) is also shown. AT 2023fhn is the faintest LFBOT in X-rays at early times. Assuming a shallow decay initially, similar to AT 2022tsd, ZTF 20acigmel and AT 2018cow, the break time can be - at the latest - similar to AT 2018cow and ZTF 20acigmel. There appears to be a correlation between break time and X-ray luminosity, with brighter LFBOTs transitioning to a steeper decay at later times. Assuming instead that epochs 1 and 2 are on the same phase of the light-curve, the decay index (where ). Expectations for the X-ray decay rate are (shock power), (magnetar central engine) and (fallback, i.e. a TDE). Overall, the detections and upper-limits are consistent with AT 2023fhn behaving like a fainter version of previous LFBOTs in the X-ray band, and demonstrates that they can exhibit several orders of magnitude of variety in their X-ray luminosity.
4.3 UV/X-ray ratio
Motivated by the fact that AT 2023fhn appears to be the brightest LFBOT yet at UV-optical wavelengths, and the faintest in terms of X-ray luminosity, in Figure 8 we show the ratio of X-ray to UV luminosity for the 3 LFBOTs with such constraints. The data points for AT 2023fhn take the X-ray detections at 12 and 23 rest-frame days, and the corresponding point on the shifted AT 2018cow light-curve in Figure 6. The uncertainties shown are exclusively from the X-ray observations. For AT 2018cow, we take the ratio of the X-ray fit of Migliori et al. (2024) in Figure 7, and our fit to the UV light-curve fit in Figure 6. Finally, for ZTF 20acigmel we take the ratio of the X-ray luminosity with the UV light-curve fit at the same time. LFBOTs therefore exhibit at least orders of magnitude in their X-ray/UV luminosity ratio, even at similar times in their evolution. This is plausibly a viewing angle effect. A qualitative prediction of tidal disruption models is a trade-off between UV-optical and X-ray luminosity as a function of viewing angle, where on-axis angles (which may also be aligned with a beamed outflow) would see a higher X-ray luminosity (Dai et al., 2018; Hayasaki & Jonker, 2021). Differences in LX/LUV are also expected for different black hole masses and spins, due to varying accretion disc formation rates (which in turn affects the delay betweeen peak X-ray and UV/optical emission, Jonker et al., 2020). However, a scenario in which the peak X-ray emission is delayed due to a delay in forming the inner accretion disc is hard to reconcile with the energetics and (variability) timescales of LFBOT emission, which demands energy input from a central engine and therefore active accretion (e.g. Ho et al., 2019; Margutti et al., 2019). Alternatively, the range of / could reflect differences in the circumstellar media, which we investigate in the following Section.
4.4 Radio
Assuming that the radio emission is synchrotron-dominated with self-absorption - as we will see, the radio SED of AT 2023fhn is consistent with this - and that the peak of the SED occurs at the synchtrotron self-absorption (SSA) frequency, we can estimate several shock parameters, and properties of the circumstellar medium. We follow the synchrotron self-absorption model of Chevalier (1998) (see also Soderberg et al., 2005). Adopting this framework for AT 2023fhn is reasonable since this best fits other LFBOTs studied so far (based on the brightness temperature, which precludes thermal emission, and the spectral shape, e.g. Margutti et al., 2019; Coppejans et al., 2020; Ho et al., 2020; Nayana & Chandra, 2021; Ho et al., 2022; Yao et al., 2022; Bright et al., 2022).
We fit the radio spectrum at 90 and 138 days (70 and 110 rest-frame days) following Chevalier (1998); Granot & Sari (2002); Chevalier & Fransson (2006). At a given time the radio SED has the form,
(1) |
where is the flux density, is the flux at the peak (break) frequency where the optically thick and thin power laws intersect, is a smoothing factor and and are spectral indices in the optically thick and thin regimes, respectively. In our case the cooling frequency lies at higher frequencies than probed by our observations (400-800 GHz), where is given by (DeMarchi et al., 2022). We therefore expect in the optically thin regime, where is the power law index of the electron energy distribution in the shock (i.e. the number of electrons with Lorentz factor goes as ).
Using the scipy curve_fit function, and working with rest-frame times and central frequencies throughout this section, we fit equation 1 to the 90 day and 138 day (observer frame) data. At 90 days we have 6 data points (5 detections, 1 upper limit, we combine the 87 and 95 day data for this epoch), and at 138 days we have 7 data points (6 detections, 1 upper limit). There are 5 parameters to fit: , , , and . The best-fit values for these parameters and their uncertainties are listed in Table 7. The optically-thin spectral index of -0.59 (138 days) yields an electron energy spectral index of 2.2, which is relatively shallow - 2.5 is expected from theory, while values closer to 3 are often measured in gamma-ray bursts, tidal disruption events and supernovae (e.g. Chevalier & Fransson, 2006; Cendes et al., 2023). Values from other LFBOTs are also in the range 2–3 (Margutti et al., 2019; Ho et al., 2020; Coppejans et al., 2020; Yao et al., 2022; Bright et al., 2022).
The peak flux Fpk and (rest-frame) frequency at the peak flux (at the intersection of the power-laws, rather than the fitted peak) allow us to estimate the radius of the shock, circumstellar density at that radius, and the CSM surface density parameter (see DeMarchi et al., 2022, for a detailed description of the modelling assumptions). Following the formulism of Chevalier (1998) (see also Chevalier & Fransson 2006; DeMarchi et al. 2022; Bright et al. 2022), we first have the shock radius , given by,
(2) |
where Dθ is the angular diameter distance, and and are the fraction of the shock energy in electrons and in the magnetic field, respectively. The average shock velocity can then be calculated as , where , is the Lorentz factor and is the rest-frame observation time. Next we have, for the internal magnetic field ,
(3) |
and for the wind density (the mass loss rate over the wind velocity),
(4) |
Under the assumption that the CSM is dominated by fully ionised hydrogen, the electron number density can be related to by - where is the proton mass - so that,
(5) |
Additionally we have, for the internal shock energy ,
(6) |
We assume equipartition (==1/3), where the magnetic energy density, the energy density in electrons and the energy density in protons contribute equally as destinations for the converted kinetic energy in the shock. We further assume for the filling factor. If the emission region is modelled as a disc of radius and thickness on the sky, whose volume is , an equivalent spherical volume can be given by . The filling factor is the fraction of this equivalent spherical volume producing emission (Chevalier, 1998).
We list the inferred properties of AT 2023fhn’s blast-wave in Table 7. Results for the fiducial parameters of and are also listed. These properties are compared with other LFBOTs in Figures 9, 10 and 11. In LFBOTs the expanding blast-wave typically shows a SSA spectrum that decreases in peak flux and frequency over time. However we note that AT2023fhn shows an increase in peak flux between and days post explosion. A similar increase was seen in CSS161010 between 69 and 99 days post explosion (Coppejans et al., 2020). This could potentially be caused by an increase in density, or inhomogeneities in the CSM, but we are not able to test this scenario given our weak constraints on the SSA peak at days post explosion.
Finally, we calculate a dimensionless normalisation of the wind density parameter (Chevalier & Li, 2000),
(7) |
where for a Wolf-Rayet-like wind with M⊙yr-1 and km s-1. From our best fits to the radio data, we derive that AT 2023fhn at 70–110 rest-frame days has ( M⊙ yr-1 for = 1000 km s-1). This mass loss rate is consistent with that of Wolf-Rayet stars. As shown in Figure 11, this density is also consistent with that of the other LFBOTs. The constraints on the synchrotron self-absorption peak at days post explosion were unfortunately insufficient to constrain the density profile of the CSM around AT2023fhn.
days | ||
Parameter | Unit | Value |
GHz | 44 | |
erg s-1 Hz-1 | (4) | |
– | 3.00.7 | |
– | -0.660.04 | |
cm | 1.30.7 | |
– | 0.70.4 | |
/vw | 10-4 M⊙ yr-1 / 1000 km s-1 | 0.08 |
cm-3 | 21 | |
G | 0.24 | |
erg | 2.2 | |
cm | 1.2 | |
– | 0.6 | |
/vw | 10-4 M⊙ yr-1 / 1000 km s-1 | 1.0 |
cm-3 | (3.4) | |
G | 0.15 | |
erg | 19 |
days | ||
Parameter | Unit | Value |
GHz | 53 | |
erg s-1 Hz-1 | (64) | |
– | 2.810.02 | |
– | -0.60.2 | |
cm | 1.2 | |
– | 0.4 | |
/vw | 10-4 M⊙ yr-1 / 1000 km s-1 | 0.32 |
cm-3 | (1.0) | |
G | 0.31 | |
erg | 2.8 | |
cm | 1.1 | |
– | 0.4 | |
/vw | 10-4 M⊙ yr-1 / 1000 km s-1 | 4.0 |
cm-3 | (1.6) | |
G | 0.19 | |
erg | 25 |
5 Discussion
In this Section, we discuss the host galaxy and derived properties of AT 2023fhn in the context of the other LFBOTs discovered thus far.
The host of AT 2023fhn, taking the spiral and satellite as one interacting system, is broadly consistent with the host galaxies of core-collapse supernovae, and slightly above four other LFBOT hosts in terms of specific star formation rate (although below the host of ZTF 18abvkwla). The host offset and faint, diffuse emission at the transient location (see Section 3.1) are consistent with the tail of the core-collapse supernova distribution (see also Chrimes et al., 2024). While the local and broader environment is consistent with a core-collapse origin, it is interesting to consider whether the high sSFR is related to tidal interaction between the spiral and satellite galaxy. Such interactions may be associated with an elevated tidal disrupted event (TDE) rate, which show a bias towards occurrence in post-starburst galaxies and galaxies undergoing interactions/mergers (French et al., 2016). The non-nuclear location of AT 2023fhn - at high offset from both the spiral and satellite - is difficult to explain in such a scenario (Chrimes et al., 2024). On the other hand, the measured optical magnitudes at the location of AT 2023fhn (see Table 4) allow a contribution from a point source with absolute magnitude as bright as -14. A significant contribution from a point source at the location of AT 2023fhn is disfavoured (as the precise location has similar brightness to its immediate surroundings, as explained in Section 3.1), but the presence of a globular cluster or ultra-compact dwarf galaxy - which may host massive black holes (e.g. Seth et al., 2014) - cannot be ruled out. The presence of such an undetected cluster or ultra compact dwarf galaxy would be consistent with the upper limit on the black hole mass inferred from our late-time UV observations of M⊙ (Mummery et al., 2024), given black hole - host galaxy/cluster mass relations (Kormendy & Ho, 2013; Lützgendorf et al., 2013).
The UV-optical, X-ray and radio evolution of AT 2023fhn is broadly similar to other LFBOTs. Notably, however, the X-ray to UV luminosity ratio of AT 2023fhn is an order of magnitude lower than AT 2018cow at similar times, and up to 3 orders of magnitude lower than ZTF 20acigmel. As we show in Section 4.4, it is difficult to attribute this variety to differences in the circumstellar medium density or blast-wave propagation, as AT 2023fhn has a blast-wave velocity, energy and CSM comparable with other LFBOTs. This is consistent with the evidence from other LFBOTs thus far that the X-ray emission arises from a distinct mechanism, namely central engine activity. AT 2023fhn is only the third LFBOT with a mildly relativistic outflow (), in common with CSS161010 and ZTF 18abvkwla, demonstrating that the blast-wave is engine-driven. As can be seen in Figure 11, all LFBOTs with sufficient constraints from radio observations thus far have a wind-like (in the sense that density decreases with distance), albeit not , circumstellar density profile. This suggests the CSM was produced by the progenitor system (i.e. through winds), rather than the explosion occurring in a pre-existing dense ISM, which would produce a flat density profile.
An alternative explanation for the variety in UV/X-ray ratios is the viewing angle, where the asymmetric outflow and accretion disc are being viewed from different angles. In this interpretation, the viewing angle to AT 2018cow was closer to perpendicular to the plane of the accretion disc (although not exactly perpendicular, Margutti et al., 2019). This conclusion was also reached by Maund et al. (2023) based on the high polarization (and possibly for AT 2022tsd, given the observation of late-time optical flares Ho et al., 2023c). AT 2023fhn, meanwhile, would have been seen close to edge-on, well off-axis from any asymmetric outflow (e.g. from a choked jet) and with the inner accretion disc obscured (where choked jets and/or the inner disc dominates the X-ray luminosity, Jonker et al., 2020). The effect of viewing angle as an explanation for different LFBOT X-ray luminosities has also been posited by Metzger (2022) and Migliori et al. (2024).
With the fundamental LFBOT requirement of a low 56Ni ejecta mass, and magnetar central engines struggling to explain all aspects of LFBOT phenomenology (e.g. the late-time emission in AT2018cow, Chen et al., 2023; Li et al., 2024), constraints on the possible progenitor models are tightening. IMBH TDE models remain plausible, if a dense wind-like CSM can be produced (e.g. by the ejection of stripped mass during the disruption event). However, the star-forming nature of the host galaxy population, and the locations of LFBOTs within them, likely disfavour such an interpretation. Other plausible models include the delayed merger of black holes and Wolf-Rayet stars (Metzger, 2022), and failed supernovae (Quataert et al., 2019). In AT 2023fhn, the mass-loss wind parameter A - higher than many radio loud supernovae (Chevalier & Fransson, 2006) and collapsar GRBs (Gompertz et al., 2018; Chrimes et al., 2022). Such a dense circumstellar environment likely require a short-lived evolutionary stage with enhanced mass loss, for example pre-explosion winds from a blue supergiant or Wolf-Rayet star (Margutti et al., 2019).
6 Conclusions
We list here our conclusions about the nature of AT 2023fhn and its place in the context of other LFBOTs and extragalactic transients more generally,
-
1.
Although relatively isolated compared with other LFBOTs and indeed most core-collapse supernovae, AT 2023fhn is otherwise consistent with a core-collapse event, associated with a typical star-forming galaxy and located in a young (albeit diffuse) stellar population.
-
2.
The low X-ray to UV luminosity ratio demonstrates orders of magnitude of variety in this parameter among LFBOTs, which may be indicative of differences in viewing angle. In this interpretation, the relatively low X-ray luminosity of AT 2023fhn is due to an edge-on viewing angle, such that the inner accretion disc is obscured and we are well off-axis from any choked jet/asymmetric outflow.
-
3.
The CSM properties are similar to previous LFBOTs, and are indicative of a dense surrounding medium. Given the wind-like ne density profiles of other LFBOTs, and our ne measurements of AT 2023fhn which continue this trend, it is likely that the dense CSM was produced by wind-like mass-loss from the progenitor system itself (rather than the progenitor exploding in a pre-existing dense ISM).
-
4.
An IMBH TDE interpretation remains possible, only if there exists a pre-existing dense CSM or if the early stages of the tidal disruption produce such an environment. The host galaxy is likely undergoing tidal interactions, which may elevate the TDE rate. While the non-nuclear location and host properties rate favour a core-collapse origin, the explosion of AT 2023fhn in an undetected globular cluster or ultra-compact dwarf galaxy cannot be ruled out.
Despite mounting evidence, the origin of LFBOTs is still ambiguous. Two approaches will elucidate which of the proposed scenarios contribute to the LFBOT population. The first is to grow the sample, enabling statistically robust comparisons of offsets and host galaxy properties to be made with other classes of transient. This will be possible with the advent of new wide-field, deep sky surveys such as those performed by the Vera Rubin observatory (Ivezić et al., 2019). The second is to perform detailed studies of future local events - like AT2018cow - across the electromagnetic spectrum. Such events offer the best opportunity to search for underlying clusters, monitor the long-term evolution, understand the detailed emission physics, and ultimately, determine their progenitors. Although there is much progress still to be made, based on AT 2023fhn and the growing population of LFBOTs, we deem a massive star progenitor with strong winds but low ejecta mass the most likely scenario. This favours models such as black hole/Wolf-Rayet mergers or failed supernovae.
Acknowledgements.
A.A.C. acknowledges support from the European Space Agency (ESA) as an ESA Research Fellow. P.G.J. has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 101095973). P.J.G. is partly supported by NRF SARChI Grant 111692.Observations analysed in this work were taken by the NASA/ESA Hubble Space Telescope under program 17238. This research has made use of software provided by the Chandra X-ray Center (CXC) in the application of the CIAO package (Fruscione et al., 2006). The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.
Computing facilities were provided by the Scientific Computing Research Technology Platform of the University of Warwick. This research has made use of the Spanish Virtual Observatory (https://svo.cab.inta-csic.es) project funded by MCIN/AEI/10.13039/501100011033/ through grant PID2020-112949GB-I00. The Pan-STARRS1 Surveys (PS1) and the PS1 public science archive have been made possible through contributions by the Institute for Astronomy, the University of Hawaii, the Pan-STARRS Project Office, the Max-Planck Society and its participating institutes, the Max Planck Institute for Astronomy, Heidelberg and the Max Planck Institute for Extraterrestrial Physics, Garching, The Johns Hopkins University, Durham University, the University of Edinburgh, the Queen’s University Belfast, the Harvard-Smithsonian Center for Astrophysics, the Las Cumbres Observatory Global Telescope Network Incorporated, the National Central University of Taiwan, the Space Telescope Science Institute, the National Aeronautics and Space Administration under Grant No. NNX08AR22G issued through the Planetary Science Division of the NASA Science Mission Directorate, the National Science Foundation Grant No. AST-1238877, the University of Maryland, Eotvos Lorand University (ELTE), the Los Alamos National Laboratory, and the Gordon and Betty Moore Foundation. This publication makes use of data products from the Wide-field Infrared Survey Explorer, which is a joint project of the University of California, Los Angeles, and the Jet Propulsion Laboratory/California Institute of Technology, funded by the National Aeronautics and Space Administration.
References
- Avni (1976) Avni, Y. 1976, ApJ, 210, 642
- Barbary (2016) Barbary, K. 2016, extinction v0.3.0
- Bright et al. (2022) Bright, J. S., Margutti, R., Matthews, D., et al. 2022, ApJ, 926, 112
- Cendes et al. (2023) Cendes, Y., Berger, E., Alexander, K. D., et al. 2023, arXiv e-prints, arXiv:2308.13595
- Chambers et al. (2016) Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, arXiv e-prints, arXiv:1612.05560
- Chen et al. (2023) Chen, Y., Drout, M. R., Piro, A. L., et al. 2023, ApJ, 955, 43
- Chevalier (1998) Chevalier, R. A. 1998, ApJ, 499, 810
- Chevalier & Fransson (2006) Chevalier, R. A. & Fransson, C. 2006, ApJ, 651, 381
- Chevalier & Li (2000) Chevalier, R. A. & Li, Z.-Y. 2000, ApJ, 536, 195
- Chrimes et al. (2022) Chrimes, A. A., Gompertz, B. P., Kann, D. A., et al. 2022, MNRAS, 515, 2591
- Chrimes et al. (2024) Chrimes, A. A., Jonker, P. G., Levan, A. J., et al. 2024, MNRAS, 527, L47
- Conroy & Gunn (2010) Conroy, C. & Gunn, J. E. 2010, ApJ, 712, 833
- Conroy et al. (2009) Conroy, C., Gunn, J. E., & White, M. 2009, ApJ, 699, 486
- Conselice (2003) Conselice, C. J. 2003, ApJS, 147, 1
- Coppejans et al. (2020) Coppejans, D. L., Margutti, R., Terreran, G., et al. 2020, ApJ, 895, L23
- Dai et al. (2018) Dai, L., McKinney, J. C., Roth, N., Ramirez-Ruiz, E., & Miller, M. C. 2018, ApJ, 859, L20
- DeMarchi et al. (2022) DeMarchi, L., Margutti, R., Dittman, J., et al. 2022, ApJ, 938, 84
- Dickey & Lockman (1990) Dickey, J. M. & Lockman, F. J. 1990, ARA&A, 28, 215
- DuPont et al. (2022) DuPont, M., MacFadyen, A., & Zrake, J. 2022, ApJ, 931, L16
- Eldridge et al. (2017) Eldridge, J. J., Stanway, E. R., Xiao, L., et al. 2017, PASA, 34, e058
- Fitzpatrick (1999) Fitzpatrick, E. L. 1999, PASP, 111, 63
- Foreman-Mackey (2016) Foreman-Mackey, D. 2016, The Journal of Open Source Software, 1, 24
- Foreman-Mackey et al. (2013) Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, 306
- Fox & Smith (2019) Fox, O. D. & Smith, N. 2019, MNRAS, 488, 3772
- French et al. (2016) French, K. D., Arcavi, I., & Zabludoff, A. 2016, ApJ, 818, L21
- Fruscione et al. (2006) Fruscione, A., McDowell, J. C., Allen, G. E., et al. 2006, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 6270, Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, ed. D. R. Silva & R. E. Doxsey, 62701V
- Gallazzi et al. (2005) Gallazzi, A., Charlot, S., Brinchmann, J., White, S. D. M., & Tremonti, C. A. 2005, MNRAS, 362, 41
- Gompertz et al. (2018) Gompertz, B. P., Fruchter, A. S., & Pe’er, A. 2018, ApJ, 866, 162
- Gottlieb et al. (2022) Gottlieb, O., Tchekhovskoy, A., & Margutti, R. 2022, MNRAS, 513, 3810
- Granot & Sari (2002) Granot, J. & Sari, R. 2002, ApJ, 568, 820
- Hayasaki & Jonker (2021) Hayasaki, K. & Jonker, P. G. 2021, ApJ, 921, 20
- Ho (2023) Ho, A. Y. Q. 2023, Transient Name Server AstroNote, 174, 1
- Ho et al. (2023a) Ho, A. Y. Q., Bremer, M., Schulze, S., & Perley, D. 2023a, Transient Name Server AstroNote, 100, 1
- Ho et al. (2023b) Ho, A. Y. Q., Liu, C., Andreoni, I., et al. 2023b, Transient Name Server AstroNote, 93, 1
- Ho et al. (2022) Ho, A. Y. Q., Margalit, B., Bremer, M., et al. 2022, ApJ, 932, 116
- Ho et al. (2023c) Ho, A. Y. Q., Perley, D. A., Chen, P., et al. 2023c, Nature, 623, 927
- Ho et al. (2023d) Ho, A. Y. Q., Perley, D. A., Gal-Yam, A., et al. 2023d, ApJ, 949, 120
- Ho et al. (2020) Ho, A. Y. Q., Perley, D. A., Kulkarni, S. R., et al. 2020, ApJ, 895, 49
- Ho et al. (2019) Ho, A. Y. Q., Phinney, E. S., Ravi, V., et al. 2019, ApJ, 871, 73
- Hoffmann et al. (2021) Hoffmann, S. L., Mack, J., Avila, R., et al. 2021, in American Astronomical Society Meeting Abstracts, Vol. 53, American Astronomical Society Meeting Abstracts, 216.02
- Inkenhaag et al. (2023) Inkenhaag, A., Jonker, P. G., Levan, A. J., et al. 2023, MNRAS, 525, 4042
- Ivezić et al. (2019) Ivezić, Ž., Kahn, S. M., Tyson, J. A., et al. 2019, ApJ, 873, 111
- Johnson et al. (2023) Johnson, B., Foreman-Mackey, D., Sick, J., et al. 2023, dfm/python-fsps: v0.4.6
- Johnson et al. (2021) Johnson, B. D., Leja, J., Conroy, C., & Speagle, J. S. 2021, ApJS, 254, 22
- Jonker et al. (2020) Jonker, P. G., Stone, N. C., Generozov, A., van Velzen, S., & Metzger, B. 2020, ApJ, 889, 166
- Kelly & Kirshner (2012) Kelly, P. L. & Kirshner, R. P. 2012, ApJ, 759, 107
- Khatami & Kasen (2023) Khatami, D. & Kasen, D. 2023, arXiv e-prints, arXiv:2304.03360
- Kormendy & Ho (2013) Kormendy, J. & Ho, L. C. 2013, ARA&A, 51, 511
- Kron (1980) Kron, R. G. 1980, ApJS, 43, 305
- Kuin et al. (2019) Kuin, N. P. M., Wu, K., Oates, S., et al. 2019, MNRAS, 487, 2505
- Leja et al. (2017) Leja, J., Johnson, B. D., Conroy, C., van Dokkum, P. G., & Byler, N. 2017, ApJ, 837, 170
- Li et al. (2024) Li, L., Zhong, S.-Q., Xiao, D., et al. 2024, ApJ, 963, L13
- Lützgendorf et al. (2013) Lützgendorf, N., Kissler-Patig, M., Neumayer, N., et al. 2013, A&A, 555, A26
- Lyutikov & Toonen (2019) Lyutikov, M. & Toonen, S. 2019, MNRAS, 487, 5618
- Margutti et al. (2019) Margutti, R., Metzger, B. D., Chornock, R., et al. 2019, ApJ, 872, 18
- Martin et al. (2003) Martin, C., Barlow, T., Barnhart, W., et al. 2003, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 4854, Future EUV/UV and Visible Space Astrophysics Missions and Instrumentation., ed. J. C. Blades & O. H. W. Siegmund, 336–350
- Matthews et al. (2023) Matthews, D., Margutti, R., Metzger, B. D., et al. 2023, Research Notes of the American Astronomical Society, 7, 126
- Maund et al. (2023) Maund, J. R., Höflich, P. A., Steele, I. A., et al. 2023, MNRAS, 521, 3323
- Metzger (2022) Metzger, B. D. 2022, ApJ, 932, 84
- Metzger & Perley (2023) Metzger, B. D. & Perley, D. A. 2023, ApJ, 944, 74
- Migliori et al. (2024) Migliori, G., Margutti, R., Metzger, B. D., et al. 2024, ApJ, 963, L24
- Mummery et al. (2024) Mummery, A., van Velzen, S., Nathan, E., et al. 2024, MNRAS, 527, 2452
- Nayana & Chandra (2021) Nayana, A. J. & Chandra, P. 2021, ApJ, 912, L9
- Oke & Gunn (1982) Oke, J. B. & Gunn, J. E. 1982, PASP, 94, 586
- Pasham et al. (2021) Pasham, D. R., Ho, W. C. G., Alston, W., et al. 2021, Nature Astronomy, 6, 249
- Pellegrino et al. (2022) Pellegrino, C., Howell, D. A., Vinkó, J., et al. 2022, ApJ, 926, 125
- Perley et al. (2021) Perley, D. A., Ho, A. Y. Q., Yao, Y., et al. 2021, MNRAS, 508, 5138
- Perley et al. (2019) Perley, D. A., Mazzali, P. A., Yan, L., et al. 2019, MNRAS, 484, 1031
- Petrosian (1976) Petrosian, V. 1976, ApJ, 210, L53
- Prentice et al. (2018) Prentice, S. J., Maguire, K., Smartt, S. J., et al. 2018, ApJ, 865, L3
- Quataert et al. (2019) Quataert, E., Lecoanet, D., & Coughlin, E. R. 2019, MNRAS, 485, L83
- Rivera Sandoval et al. (2018) Rivera Sandoval, L. E., Maccarone, T. J., Corsi, A., et al. 2018, MNRAS, 480, L146
- Rodrigo & Solano (2020) Rodrigo, C. & Solano, E. 2020, in XIV.0 Scientific Meeting (virtual) of the Spanish Astronomical Society, 182
- Rodrigo et al. (2012) Rodrigo, C., Solano, E., & Bayo, A. 2012, SVO Filter Profile Service Version 1.0, IVOA Working Draft 15 October 2012
- Rodriguez-Gomez et al. (2019) Rodriguez-Gomez, V., Snyder, G. F., Lotz, J. M., et al. 2019, MNRAS, 483, 4140
- Schlafly & Finkbeiner (2011) Schlafly, E. F. & Finkbeiner, D. P. 2011, ApJ, 737, 103
- Schrøder et al. (2020) Schrøder, S. L., MacLeod, M., Loeb, A., Vigna-Gómez, A., & Mandel, I. 2020, ApJ, 892, 13
- Schulze et al. (2021) Schulze, S., Yaron, O., Sollerman, J., et al. 2021, ApJS, 255, 29
- Seth et al. (2014) Seth, A. C., van den Bosch, R., Mieske, S., et al. 2014, Nature, 513, 398
- Soderberg et al. (2005) Soderberg, A. M., Kulkarni, S. R., Berger, E., et al. 2005, ApJ, 621, 908
- Soker (2022) Soker, N. 2022, Research in Astronomy and Astrophysics, 22, 055010
- Stanway & Eldridge (2018) Stanway, E. R. & Eldridge, J. J. 2018, MNRAS, 479, 75
- Sun et al. (2022) Sun, N.-C., Maund, J. R., Crowther, P. A., & Liu, L.-D. 2022, MNRAS, 512, L66
- Sun et al. (2023) Sun, N.-C., Maund, J. R., Shao, Y., & Janiak, I. A. 2023, MNRAS, 519, 3785
- Tonry et al. (2012) Tonry, J. L., Stubbs, C. W., Lykke, K. R., et al. 2012, ApJ, 750, 99
- Tremonti et al. (2004) Tremonti, C. A., Heckman, T. M., Kauffmann, G., et al. 2004, ApJ, 613, 898
- Uno & Maeda (2020) Uno, K. & Maeda, K. 2020, ApJ, 897, 156
- Xiang et al. (2021) Xiang, D., Wang, X., Lin, W., et al. 2021, ApJ, 910, 42
- Yao et al. (2022) Yao, Y., Ho, A. Y. Q., Medvedev, P., et al. 2022, ApJ, 934, 104
- Zhang et al. (2022) Zhang, W., Shu, X., Chen, J.-H., et al. 2022, Research in Astronomy and Astrophysics, 22, 125016
Appendix A SED fitting MCMC results
In this appendix we provide the joint posterior parameter distributions for the host galaxy of AT 2023fhn, in the form of a corner plot (Figure 12) including stellar mass, metallicity, extinction, population age and timescale for an exponentially declining star formation history. These are provided as outputs from emcee SED-fitting using prospector. For the MCMC initial values (and flat priors) we use A (0¡AV¡2), t Gyr and M M⊙ (¡M/M⊙¡), with a flat prior on of 0.1¡/Gyr¡100. For our fiducial run, the redshift is fixed at and the metallicity at . Also provided in Table 8 are results when is allowed to vary a free parameter.
Host property | Value |
---|---|
M⋆ / M⊙ | |
SFR / M⊙ yr-1 | 7.3 |
Z/Z⊙ | 0.08 |
AV | 0.40 |
tage/Gyr | 1.8 |
/Gyr | 1.1 |