ASASSN-14ko is a Periodic Nuclear Transient in ESO 253-G003
Anna V. Payne, Benjamin J. Shappee, Jason T. Hinkle, Patrick J. Vallely, Christopher S. Kochanek, Thomas W.-S. Holoien, Katie Auchettl, K. Z. Stanek, Todd A. Thompson, Jack M. M. Neustadt, Michael A. Tucker, James D. Armstrong, Joseph Brimacombe, Paulo Cacella, Robert Cornect, Larry Denneau, Michael M. Fausnaugh, Heather Flewelling, Dirk Grupe, A.N. Heinze, Laura A. Lopez, Berto Monard, Jose L. Prieto, Adam C. Schneider, Scott S. Sheppard, John L. Tonry, Henry Weiland
DDraft version September 9, 2020
Typeset using L A TEX twocolumn style in AASTeX63
ASASSN-14ko is a Periodic Nuclear Transient in ESO 253 − G003
Anna V. Payne , ∗ Benjamin J. Shappee , Jason T. Hinkle , Patrick J. Vallely , † Christopher S. Kochanek ,
2, 3
Thomas W.-S. Holoien , ‡ Katie Auchettl ,
5, 6, 7, 8
K. Z. Stanek,
2, 3
Todd A. Thompson ,
2, 3
Jack M. M. Neustadt , Michael A. Tucker , § James D. Armstrong, Joseph Brimacombe, Paulo Cacella, Robert Cornect, Larry Denneau, Michael M. Fausnaugh , Heather Flewelling , Dirk Grupe , A. N. Heinze , Laura A. Lopez,
2, 3
Berto Monard, Jose L. Prieto ,
16, 17
Adam C. Schneider ,
18, 19
Scott S. Sheppard , John L. Tonry, and Henry Weiland Institute for Astronomy, University of Hawai`i at Manoa, 2680 Woodlawn Dr., Honolulu, HI 96822 Department of Astronomy, The Ohio State University, 140 West 18th Avenue, Columbus, OH 43210, USA Center for Cosmology and AstroParticle Physics, The Ohio State University, 191 W. Woodruff Ave., Columbus, OH 43210, USA The Observatories of the Carnegie Institution for Science, 813 Santa Barbara St., Pasadena, CA 91101, USA School of Physics, The University of Melbourne, Parkville, VIC 3010, Australia ARC Centre of Excellence for All Sky Astrophysics in 3 Dimensions (ASTRO 3D) Department of Astronomy and Astrophysics, University of California, Santa Cruz, CA 95064, USA DARK, Niels Bohr Institute, University of Copenhagen, Lyngbyvej 2, 2100 Copenhagen, Denmark Institute for Astronomy, University of Hawai`i, 34 Ohia Ku St., Pukalani, HI 96768, USA Coral Towers Observatory, Cairns, QLD 4870, Australia Dogsheaven Observatory, SMPW Q25CJ1 LT10B, Brasilia, Brazil Moondyne Observatory, Bakers Hill, Western Australia, Australia Department of Physics, and Kavli Institute for Astrophysics and Space Research, Massachusetts Institute of Technology, Cambridge,MA 02139, USA Department of Physics, Earth Science, and Space System Engineering, Morehead State University, 235 Martindale Dr, Morehead, KY40351 Bronberg Observatory, Center for Backyard Astrophysics Pretoria, PO Box 11426, Tiegerpoort 0056, South Africa; KleinkarooObservatory, Center for Backyard Astrophysics Kleinkaroo, Sint Helena 1B, PO Box 281, Calitzdorp 6660, South Africa N´ucleo de Astronom´ıa de la Facultad de Ingenier´ıa y Ciencias, Universidad Diego Portales, Av. Ej´ercito 441, Santiago, Chile Millennium Institute of Astrophysics, Santiago, Chile US Naval Observatory, Flagstaff Station, P.O. Box 1149, Flagstaff, AZ 86002, USA Department of Physics and Astronomy, George Mason University, MS3F3, 4400 University Drive, Fairfax, VA 22030, USA Carnegie Institution for Science, Earth and Planets Laboratory, 5241 Broad Branch Road, Washington, DC 20015, USA (Received September 9, 2020)
Submitted to ApJABSTRACTWe present the discovery that ASASSN-14ko is a periodically flaring AGN at the center of the galaxyESO 253 − G003. At the time of its discovery by the All-Sky Automated Survey for Supernovae (ASAS-SN), it was classified as a supernova close to the nucleus. The subsequent six years of V - and g -bandASAS-SN observations reveal that ASASSN-14ko has nuclear flares occurring at regular intervals. Theseventeen observed outbursts show evidence of a decreasing period over time, with a mean period of P = 114 . ± . P = − . ± . ∼ TESS observed an outburst during Sectors 4-6, revealing a rise time of 5 . ± . Corresponding author: Anna V. [email protected] a r X i v : . [ a s t r o - ph . H E ] S e p Payne et al. scenarios to explain ASASSN-14ko’s periodic outbursts, but currently favor a repeated partial TDE.The next outbursts should peak in the optical on UT 2020-09-7 . ± . . ± . Keywords: black hole physics – galaxies: nucleus – Seyfert galaxy – accretion, accretion discs INTRODUCTIONThere are numerous physical processes that lead tovariability in the nuclei of galaxies. Every massivegalaxy likely houses a supermassive black hole (SMBH;Kormendy & Richstone 1995, Kormendy & Ho 2013),and the past several decades have been spent unrav-eling their accretion and variability processes (for re-views, see, e.g., Ulrich et al. 1997, Ho 2008, Heckman& Best 2014, Yuan & Narayan 2014, Padovani et al.2017, Hickox & Alexander 2018, Komossa 2018, Bland-ford et al. 2019). Without the ability to spatially re-solve the immediate vicinity of the SMBH, other meth-ods must be used to probe accretion physics.Variable accretion in active galactic nuclei (AGN) isthe primary driver of nuclear variability. Most quasarsappear to vary in brightness stochastically with statis-tical properties that can be modeled relatively well bya Damped Random Walk (DRW; e.g., Kelly et al. 2008,Koz(cid:32)lowski et al. 2010, MacLeod et al. 2010, Zu et al.2013). Signatures that deviate from DRW behavior,namely periodic or semi-periodic features, have beensuggested as possible indicators for a binary system atthe galaxy’s core (e.g., Komossa 2006).This has led to searches for periodic signals in AGNlight curves to identify SMBH binaries. For exam-ple, Graham et al. (2015a) used the Catalina Real-time Transient Survey (CRTS) to search for sub-parsecSMBH binaries. They reported 111 candidates thatshowed evidence of periodicity associated with a Ke-plerian orbit. Another study by Charisi et al. (2016)used the Palomar Transient Factory to identify 33 can-didates with evidence for periodic variability. Liu etal. (2015; 2016; 2019) searched for periodicity in Pan-STARRS1’s Medium Deep Survey, ultimately findingone candidate, PSO J185. Other candidates with quasi-periodic/periodic variability include NGC 4151 with anestimated period of ∼
16 years (e.g., Oknyanskij 1978,Pacholczyk et al. 1983, Guo et al. 2006, Oknyanskij &Lyuty 2007, Bon et al. 2012), and PG 1302-102 with anestimated period of 1,884 days (Graham et al. 2015b).Simulations have shown that periodic variability is ex- ∗ NASA Fellow † NSF Graduate Research Fellow ‡ Carnegie Fellow § DOE CSGF Fellow pected in the light curves of SMBH binaries at sub-parsec separations due to a variety of processes, in-cluding modulated mass accretion onto the binary (e.g.,D’Orazio et al. 2013, Gold et al. 2014, Farris et al. 2014)or relativistic Doppler boosting of the minidisks formedas a result of the binary interaction (D’Orazio et al.2015).Aside from low-level variability, AGN can also showoutbursts or flares in which the brightness of the AGNincreases dramatically for a short period of time beforereturning to a level of relative quiescence. The best ex-ample of quasi-periodic optical flares is the 12-year out-burst cycle of OJ 287. These were first reported bySillanpaa et al. (1988), who suggested that outburstsare due to perturbations of the primary black hole’s ac-cretion disk during pericenter passages of the secondaryblack hole on a 12 year orbital cycle. Relativistic effectslike precession probably alter the orbital geometry sothat the events are not strictly periodic (see, Valtonenet al. 2006, Laine et al. 2020). The most recent flaredetected from OJ 287 brightened in the X-ray, UV, andoptical and occurred between April-June 2020, whichwas consistent with the predictions of the binary blackhole model (Komossa et al. 2020). Another candidateis IC 3599, which has competing theories for the causeof its X-ray/optical flares. Grupe et al. (2015) proposedaccretion disk instabilites as the cause of the recurring ∼ t − / .In the case of a partial TDE, the star survives the en-counter with the SMBH and only a fraction of the stellarmaterial is tidally stripped, leaving the stellar core in-tact. Guillochon & Ramirez-Ruiz (2013) found that fall-back rate for partial TDEs at late times becomes steeperthan t − / because there is less debris with orbital bind- SASSN-14ko t − / , which is effectively indepen-dent of the mass of the core that survives the passageclose to the black hole. This fallback rate is supported bythe hydrodynamical simulations of Miles et al. (2020).Hydrodynamical simulations also indicate that partialdisruptions can repeat, causing episodic mass transferfrom the star to the SMBH at every pericenter passage,resulting in a series of low-level flares that repeat onthe orbital timescale (MacLeod et al. 2013). Partial dis-ruptions are most easily achieved for giant stars (e.g.,MacLeod et al. 2013, Guillochon & Ramirez-Ruiz 2013)which might also be created as stellar merger products(e.g., Antonini et al. 2011, MacLeod et al. 2012).Most theoretical predictions for TDEs predate any ob-servations of the phenomenon. The first observationalclaims of TDEs were soft X-ray outbursts from other-wise quiescent galaxies (e.g., Bade et al. 1996, Komossa& Greiner 1999, Grupe et al. 1999, Greiner et al. 2000,Gezari et al. 2003, Komossa 2015). Since then, TDEflares have been discovered at a range of wavelengths,including the hard X-ray (e.g., Bloom et al. 2011; Bur-rows et al. 2011; Cenko et al. 2012a; Pasham et al. 2015),soft X-ray (e.g., Komossa & Bade 1999; Donley et al.2002; Maksym et al. 2010; Saxton et al. 2012), ultravi-olet (e.g., Stern et al. 2004, Gezari et al. 2006, Gezariet al. 2008, Gezari et al. 2009), and optical (e.g., vanVelzen et al. 2011, Gezari et al. 2012, Cenko et al. 2012b,Arcavi et al. 2014, Holoien et al. 2014a, Vink´o et al.2015, Holoien et al. 2016a, Holoien et al. 2016b, Holoienet al. 2018, Holoien et al. 2019a, Holoien et al. 2019b,Hinkle et al. 2020b, Holoien et al. 2020, van Velzen et al.2020). Due to the intrinsic multi-wavelength propertiesof both TDEs and AGN, one problem is to identify char-acteristics that clearly distinguish between the two ob-jects, notably using X-rays (Auchettl et al. 2018). Thisis becoming more important with the discoveries of moreambiguous transients such as ASASSN-18el (Trakhten-brot et al. 2019, Ricci et al. 2020) and ASASSN-18jd(Neustadt et al. 2020).Here we report the discovery and long-term obser-vation of a series of periodic outbursts from ASASSN-14ko, which is associated with the AGN ESO 253 − G003(z=0.042489, Aguero et al. 1996). ESO 253 − G003was spectroscopically classified as a Type 2 Seyfert byLauberts (1982). In Section 2 we discuss the discoveryof ASASSN-14ko and the photometric and spectroscopicdata used in this analysis. In Section 3 we discuss thehost properties, and in Section 4 we discuss the lightcurve and the periodic nature of the outbursts. Thespectroscopic results are presented in Section 5, and we discuss several theoretical interpretations of these peri-odic flares in Section 6. For a flat Ω m = 0 . ≈
188 Mpc and the projectedscale is ≈ .
85 kpc / arcsec. The Galactic extinction isA V = 0 .
118 mag (Schlafly & Finkbeiner 2011). DISCOVERY AND OBSERVATIONSOn 2014-11-14.28 UT, the All-Sky Automated Sur-vey for Supernovae (ASAS-SN, Shappee et al. 2014,Kochanek et al. 2017) triggered on a nuclear transientassociated with ESO 253 − G003 at V ∼ . Swift hereafter, Gehrels et al. 2004)were taken on UT 2014-11-16, 2014-11-19, 2014-11-21, 2014-11-23, and 2014-11-27 (PI: Holoien, ToO ID:33529). These observations showed that the central re-gion of the galaxy had significantly brightened in theUV but were consistent with archival magnitudes inthe optical. These
Swift data also revealed X-ray emis-sion, with fluxes of (2.85 ± × − ergs cm − s − and(3.1 ± × − ergs cm − s − on UT 2014-11-16 andUT 2014-11-19, respectively. The X-ray spectrum wasconsistent with a highly absorbed AGN with a columndensity of ∼ cm − and a luminosity of L X ∼ × ergs s − (Holoien et al. 2014b).As part of ongoing work to examine the long-termbehavior of AGN observed by ASAS-SN, a full lightcurve of ESO 253 − G003 was extracted in February 2020.Visual examination of the light curve revealed sixteenflares spaced out roughly evenly over six years, as shownin Figure 1. The seventeenth outburst in Figure 1 wasthen predicted and observed. The original ASASSN-14ko trigger corresponds to the second outburst in theseries. This initiated the further analysis and photomet-ric and spectroscopic follow-up of ASASSN-14ko whichwe report here. All photometric data used in this anal-ysis are presented in Table 1.2.1.
ASAS-SN Photometry
ASAS-SN is a network of 20 robotic telescopeshosted by Las Cumbres Observatory Global Telescope
Payne et al. F l u x [ e r g ss c m ] Fit peak: 911.4 1000 1025 1050 1075Fit peak: 1029.4 1125 1150 1175 1200Fit peak: 1138.41225 1250 1275 13002.02.53.0 Fit peak: 1246.2 1350 1375 1400 1425Fit peak: 1360.4 1450 1475 1500 15251575 1600 1625 16502.02.53.0 Outburst observed by TESSFit peak: 1585.7 1675 1700 1725 1750Fit peak: 1697.1 1800 1825 1850 18751900 1925 1950 19752.02.53.0 Fit peak: 1918.5 2000 2025 2050 2075
JD - 2456850
Fit peak: 2028.4 2125 2150 2175 2200May 2020 OutburstFit peak: 2137.5
Figure 1.
Light curves of ASASSN-14ko spanning 2014-2020 and demonstrating its periodic flaring behavior. The ASAS-SN V - and g -band data are shown in green and blue, respectively. The three TESS sector observations are included in light red.ATLAS o -band data are shown in orange and ATLAS c -band data are shown in cyan. Swift epochs are denoted by dark bluemarks, and spectral epochs are shown by orange marks at the bottom of each panel. The fifth-order polynomial fits for eachoutburst are shown in red along with shaded red regions corresponding to estimates for each peak time and its uncertainty. Themagenta vertical lines show the predicted peaks for the model with a period derivative described in Section 4.1.
SASSN-14ko Table 1.
Photometry of ASASSN-14ko used in this analysis.Only the first observation in each band is shown here todemonstrate its form and content. Table to be published inits entirety in machine-readable form in the online journal.JD Band Flux (10 − ergs s − cm − ) Flux Error (10 − ergs s − cm − )2458957.425 X-ray 0.18 0.032458957.428 UV W
UV M
UV W U B g V c o I TESS r (LCOGT, Brown et al. 2013) at five sites around theglobe. Each telescope consists of four 14-cm apertureNikon telephoto lenses with 8 . (cid:48)(cid:48) ◦ × ◦ field of view. ASAS-SN’s primary objective is todiscover supernovae by surveying the entire visible skyevery night. The ASAS-SN data shown in Figure 1 in-cludes both V -band and g -band observations. In 2018,the first two ASAS-SN mounts transitioned from V -band to g -band to match the three ASAS-SN telescopesdeployed in 2017-2018.The data were reduced using a fully-automatedpipeline based on the ISIS image subtraction package(Alard & Lupton 1998; Alard 2000). Each photomet-ric epoch (usually) combines three dithered 90-secondimage exposures with a 4.47 × apphot to perform aperture pho-tometry with a 2-pixel, or approximately 16 . (cid:48)(cid:48)
0, radiusaperture on each subtracted image, generating a dif-ferential light curve. The photometry was calibratedusing the AAVSO Photometric All-Sky Survey (Hen-den et al. 2015). All low-quality ASAS-SN images ofESO 253 − G003 were inspected by-eye, and images withclouds or other systematic problems were removed.2.2.
Swift UVOT Photometry
Following the original discovery, we requested
Swift
UltraViolet/Optical Telescope (UVOT, Roming et al.2005) ToO observations (ToO ID: 33529). Then, afterwe discovered its periodic nature, we again requested
Swift data (ToO IDs: 13836, 13979, 14005) to moni-tor ASASSN-14ko during quiescence and then duringthe outburst predicted for UT 2020-05-18.5 (see below). Data were obtained in six filters (Poole et al. 2008): V (5468 ˚A), B (4392 ˚A), U (3465 ˚A), U V W
U V M
U V W uvotsource to extract the source counts using a 16 . (cid:48)(cid:48) ∼ . (cid:48)(cid:48) Swift
UVOT B and V magnitudes to Johnson B and V magnitudes using the standard color corrections .There are three additional observations of this galaxyin the Swift data archive under the identification SWIFTJ0525 . − Swift
Burst Alert Telescope (BAT) source reported by Baum-gartner et al. (2013) in their 70 month catalog and iden-tified with the nearby blazar
P KS − − G003was in fact the BAT source rather than PKS 0524 − Swift epochs are consistent with the quiescentmagnitudes we measure from the later data and therewere too few observations to usefully add to the con-straints on the times of the outbursts.2.3.
ATLAS Photometry
The ATLAS survey (Tonry et al. 2018) consists of two0.5m f/2 Wright Schmidt telescopes on Haleakal¯a and atthe Mauna Loa Observatory. Designed primarily for de-tecting hazardous asteroids, the telescopes obtain four30-second exposures of 200-250 fields per night. Thiscorresponds to roughly a quarter of the visible sky. AT-LAS uses two broad-band filters, the ‘cyan’ ( c ) filtercovering 420 - 650 nm and the ‘orange’ ( o ) filter cover-ing 560 - 820 nm (Tonry et al. 2018).The ATLAS pipeline performs flat-field corrections foreach image as well as astrometric and photometric cali-brations. Reference images of the host galaxy were cre-ated by stacking multiple images taken under ideal con-ditions and this reference was then subtracted from eachATLAS epoch to isolate the flux from the transient. Weperformed forced photometry on the subtracted ATLAS https://heasarc.gsfc.nasa.gov/docs/heasarc/caldb/swift/docs/uvot/uvot caldb coltrans 02b.pdf Payne et al.
Table 2.
Host magnitudes of ESO 253 − G003 measured us-ing a 16 . (cid:48)(cid:48) UV W
UV M
UV W
1- and U -band magnitudes were determined from Swift data takenin April 2020 during pre-outburst quiescence. Johnson-Cousins BV R and SDSS gri magnitudes were determinedfrom LCOGT, Swift B and V , and amateur astronomer dataalso taken in April 2020. The magnitudes were combined byaveraging the data weighted by the inverse squares of theuncertainties. All magnitudes are in the AB system.Filter Magnitude Uncertainty UV W
UV M
UV W U B g V r R i J H K s W W W W images of ASASSN-14ko as described in Tonry et al.(2018). ATLAS images taken on the same night werestacked. The resulting ATLAS o - and c - band photom-etry and 3-sigma limits are included also in Figure 1.2.4. TESS Photometry
ASASSN-14ko was observed by the Transiting Exo-planet Survey Satellite (
TESS , Ricker et al. 2014) dur-ing Sectors 4-6, which occurred between 2018-10-18 and2019-01-07. Similar to the process applied to the ASAS-SN data, we used the ISIS package (Alard & Lupton1998; Alard 2000) to perform image subtraction on the30-minute cadence
TESS full frame images (FFIs) toobtain high fidelity light curves of this galaxy. This pro-cess is fully described in Vallely et al. (2019).We construct independent reference images for eachsector as opposed to utilizing a single reference imageover all sectors to avoid introducing problems createdby the field rotations between sectors. The referenceimages were built using the first 100 good-quality FFIsfor each sector. FFIs were considered poor-quality ifthe sky background levels or PSF widths were aboveaverage for the sector. FFIs were also excluded fromour analysis if they had data quality flags, or were taken when the spacecraft’s pointing was compromised due toinstrument anomalies, or when scattered light affectedthe images.The measured fluxes were converted into
TESS -bandmagnitudes using an instrumental zero point of 20.44electrons per second from the
TESS
Instrument Hand-book (Vanderspek et al. 2018).
TESS observes in a sin-gle broad-band filter, spanning roughly 6000–10000 ˚Awith an effective wavelength of ∼ TESS magnitudes are calibrated to the Vega system (Sullivanet al. 2015). The
TESS light curve is also shown inFigure 1.2.5.
Las Cumbres Observatory Global TelescopePhotometry
We obtained photometric observations from Las Cum-bres Observatory Global Telescope (LCOGT, Brownet al. 2013). The B -, V -, g (cid:48) -, r (cid:48) -, and i (cid:48) -band obser-vations were taken using the 1-meter telescope at Sid-ing Spring Observatory in New South Wales, Australia.The LCOGT photometric observations began on 2020-04-13 in quiescence and continued through the midpointof the May 2020 outburst when telescope horizon ob-serving limits prevented further observations. Aperturemagnitudes were obtained using a 16 . (cid:48)(cid:48) apphot package using an annulus to es-timate and subtract background counts. We used starswith APASS DR 10 magnitudes to calibrate the data.Similar to the process for the UVOT observations, theaperture magnitudes were corrected for Galactic extinc-tion. The host galaxy flux was measured using the April2020 quiescent data and subtracted to isolate the fluxfrom the transient.2.6. Amateur Astronomer Photometry
Amateur astronomers at four different observatoriesobserved ASASSN-14ko starting shortly after it was dis-covered to be periodically flaring. Data were taken atMoondyne Observatory, east of Perth, Australia, usinga 0.4-m telescope with AOX adaptive optics betweenApril 28 and June 8, 2020 and on a daily cadence be-tween May 10 and May 28, 2020. B C -, V -, G S -, R S -,and I C -band images were obtained with guided 120 sec-ond exposures. The data were reduced and calibratedwith standard procedures, and then stacked with 3 and5 image sets aligned using background stars. Data werealso collected using a 41-cm telescope at Savannah SkiesObservatory from Queensland, Australia. The bias anddark subtracted data were taken in the B -, V -, R C -,and I C -bands using 180 second exposures. ASASSN-14ko was observed from Bronberg Observatory in SouthAfrica using 14- and 12-inch telescopes in the R band SASSN-14ko B , V , R C , and I C fil-ters. The images were observed with 120 second expo-sures using a 14-inch telescope and calibrated.All these images were then astrometrically calibratedand aperture magnitudes were measured using the IRAF apphot package and a 16 . (cid:48)(cid:48) Swift
UVOT and LCOGTdata. 2.7.
Spectroscopic Observations
The first available observation of ESO 253 − G003 wasobtained by Kewley et al. (2001) on 1996-02-19. Atthat time, ESO 253 − G003 was classified as a Seyfert2 galaxy. When ASASSN-14ko was first discovered, wetook a follow-up spectrum using the B&C spectrographon the du Pont 2.5-m at Las Campanas Observatoryon 2014-11-16. Other spectra were taken by PESSTO(Smartt 2015) as part of transient follow-up using theESO-NTT/EFOSC2-NTT on 2014-11-25, 2014-12-12,2014-12-28, 2015-01-26, and 2015-01-27. These spectraare available at WISEREP (Yaron & Gal-Yam 2012)and in the ESO archive . Both spectra taken in Novem-ber 2014 showed noticeably broadened Balmer emission-lines compared to the archival spectrum from 1996.We also obtained seven spectra with the LCOGTFLOYDS spectrograph (Sand 2014) at the robotic 2-mFaulkes Telescope South located at Siding Spring Ob-servatory (Brown et al. 2013). These observations weretaken on 2020-04-12, 2020-04-15, 2020-04-24, 2020-04-25, 2020-04-27, 2020-05-16, and 2020-05-17 in order toobserve any changes in spectral features prior to andduring the most recent outburst. All spectra were re-duced following standard reduction procedures usingIRAF. All observations were taken with an exposuretime of 600 seconds and span 4,300 to 10,000˚A.We used the analysis tool mapspec (MCMC Algo-rithm for Parameters of Spectra; Fausnaugh et al. 2016)to calibrate the longslit spectra onto the same absoluteflux scale using the [OIII] λ mapspec uses MCMCmethods to adjust the flux, wavelength shift, and reso-lution of each individual spectrum to match that of thereference spectrum. The reference spectrum was definedby an average of the spectra sample. http://archive.eso.org https://github.com/mmfausnaugh/mapspec X-ray data
In addition to the
Swift
UVOT observations, we alsoobtained simultaneous
Swift
X-Ray Telescope (XRT,Burrows et al. 2005) photon-counting observations ofASASSN-14ko. All observations were reprocessed fromlevel one XRT data using the
Swift xrtpipeline ver-sion 0.13.2, producing cleaned event files and exposuremaps. Standard filter and screening criteria were used,as well as the most up-to date calibration files.To extract background-subtracted count rates, weused a source region with a radius of 50 (cid:48)(cid:48) centered on theoptical position of ASASSN-14ko. To define the back-ground, we used a 150 . (cid:48)(cid:48) α , δ )=(05 h m . s , − ◦ (cid:48)(cid:48) . (cid:48) ). All countrates are aperture corrected.To improve the signal to noise, we merged the mostrecent Swift
XRT observations (ObsIDs: 14005, 13979,13836) using the HEASOFT tool xselect . From thismerged observation, we extracted spectra using the task xrtproducts version 0.4.2 and the same extraction re-gions. Ancillary response files were obtained by merg-ing the individual exposure maps using
XIMAGE ver-sion 4.5.1 and the task xrtmkarf . We used the ready-made response matrix files that are available with the
Swift calibration files. This merged
Swift spectrum wasgrouped to have a minimum of 10 counts per energy binusing the
FTOOLS command grppha .On 2015 August 19, ESO 253 − G003 was observed us-ing the MOS and PN detectors onboard
XMM-Newton (ObsID: 0762920501, PI: Koss) as part of a program tostudy heavily obscured AGN. Both of the detectors wereoperated in full-frame mode using a thin filter. We re-duced the data using
XMM-Newton science system ver-sion 15.0.02 and the most up to date calibration files.Periods of high background/proton flares that could af-fect the quality of the data were identified by generatinga count rate histogram of events between 10 and 12 keV.The observations were only marginally affected by back-ground flares, leading to effective exposures of 25.6 ksand 24.0 ks for the MOS and PN detectors, respectively.For our analysis, we used standard event screening andflags for both the MOS and PN detectors . All files werecorrected for vignetting using EVIGWEIGHT . Spectrawere extracted from both detectors using the SAS taskEVSELECT and the cleaned event files. We used thesame source region used to analyze the
Swift observa- http://swift.gsfc.nasa.gov/analysis/xrt swguide v1 2.pdf https://xmm-tools.cosmos.esa.int/external/xmm user support/documentation/sas usg/USG.pdf Payne et al. tions, and the spectra were grouped using grppha tohave a minimum of 20 counts per energy bin.ESO 253 − G003 was also observed using the NuclearSpectroscopic Telescope Array (
NuSTAR ) on 2015 Au-gust 21 (ObsID: 60101014002, PI: Koss) as part ofthe same program to observe heavily obscured AGN.We reduced the data using the
NuSTAR
Data Anal-ysis Software (NuSTARDAS) Version 1.8.0 and
NuS-TAR
CALDB Version 20170817. We performed thestandard pipeline data processing with nupipeline , withthe saamode=STRICT to identify the South AtlanticAnomaly passages. Using the nuproducts
FTOOL, weextracted source spectra from a 100 (cid:48)(cid:48) -radius region andproduced ancillary response files and redistribution ma-trix files for both the A and B modules. We extractedbackground spectra from annular regions centered onthe source, and we followed the procedure outlined byWik et al. (2014) to estimate the backgrounds and sub-tract them from the source spectra using the nuskybgd routines .To analyze the merged Swift , XMM-Newton and
NuS-TAR spectra we used the X-ray spectral fitting package(XSPEC) version 12.10.1f and χ statistics. Finally, tofurther constrain the X-ray emission, we also analysedthe available XMM-Newton slew observations of the re-gion.
XMM-Newton slew observations take advantageof the fast read out of the PN detector and its abilityto observe the sky without reduction of image qualityas
XMM-Newton maneuvers between pointed observa-tions. Slew observation can detect X-ray emission downto a 0.2-10.0 keV flux limit of ∼ − ergs cm − s − (Saxton et al. 2008). We found nine slew observationsoverlapping the source and analyzed them using theSAS tool eslewchain . To extract count rates, we usethe same source and background regions used for thepointed XMM-Newton observation. Due to the low ex-posure times of each observation no spectra could beextracted. HOST PROPERTIES AND SMBH MASSWe fit the quiescent host photometry given in Table2 using
AGNfitter (Calistro Rivera et al. 2016). In ad-dition to our photometry from Section 2, we also in-clude the J , H , and K s fluxes from the 2MASS All-SkyPoint Source Catalog (Skrutskie et al. 2006), the W W µm fluxes fromthe IRAS Faint Source Catalog (Moshir & et al. 1990). https://github.com/NuSTAR/nuskybgd AGNfitter uses MCMC methods to model the combinedcontribution of the AGN accretion disk, dusty torus,stellar population, and cold dust. Based on our fit, ESO253 − G003 has a stellar mass of M ∗ = (3 . +0 . − . ) × M (cid:12) , a stellar population age of 0 . ± .
10 Gyr, and astar formation rate of SFR = 24 . ± . (cid:12) yr − .We used several approaches to estimate the SMBHmass. First, we used the 2014-11-25 PESSTO spec-trum to estimate a broad-line region radius (R BLR )using the scaling relation of Kaspi et al. (2005). Wethen fit the broad component of H β with a Gaussianto estimate the characteristic velocity of the BLR. As-suming a Keplerian orbit and using a geometry factorof 4.3 (Bentz & Katz 2015), this yields a SMBH masslog ( M BH ) = 8 . +0 . − . M (cid:12) . Second, we applied thescaling relation between bulge near-infrared (NIR) lu-minosity and SMBH mass of Marconi & Hunt (2003).Assuming the AGN contributes 33% of the total flux inthe NIR (the median value in Burtscher et al. 2015), weused archival 2MASS photometry to estimate a SMBHmass of log ( M BH ) = 7 . +0 . − . M (cid:12) , log ( M BH ) =7 . +0 . − . M (cid:12) , and log ( M BH ) = 7 . +0 . − . M (cid:12) for the J , H , and K s bands respectively. Finally, we used thehost stellar mass from our SED fits to estimate a bulgemass following Mendel et al. (2014). Then, from theM B - M BH relation of McConnell & Ma (2013), we es-timated a black hole mass of log ( M BH ) = 7 . +0 . − . M (cid:12) . A weighted average of these measurements withan uncertainty which incorporates the range of the esti-mates yields log ( M BH ) = 7 . +0 . − . M (cid:12) . We assumethis value throughout this manuscript. The correspond-ing Eddington luminosity for a black hole of this massis log ( L Edd ) = 45 . +0 . − . ergs s − . LIGHT CURVE ANALYSIS4.1.
Periodic Outbursts in the Light Curve
We individually fit each outburst with a fifth-orderpolynomial to determine the timings of the peaks asshown in red in Figure 1. We then measured the er-rors on the peak times by bootstrap resampling the lightcurves. These errors are shown by the shaded red regionsin Figure 1. The times and fluxes for each peak in the V - and g -band ASAS-SN light curve are given in Table3. Figure 1 visually demonstrates that the outburstpeaks recur at consistent intervals. Initially, we an-alyzed the light curves using the box least squares(BLS) periodogram method (Kov´acs et al. 2002) to ob-tain a period of 111.82 ± SASSN-14ko O b s e r v e d - C a l c u l a t e d P e a k s ( D a y s ) Figure 2.
O-C plot comparing the observed peak time for each outburst with the estimated peak if we assume a constantperiod. The parabolas show the predictions from the models including a ˙ P either with the assumed (solid) or re-scaled (dashed)uncertainties for the times of peak. mate was used to predict and plan for the 2020-05-18.5outburst.Using this mean period leads to significant residu-als between the observed and calculated peak times, asshown in the Observed - Calculated (O-C) diagram ofFigure 2. Such a periodic trend is indicative of a periodderivative, so for the present analysis we fit the time ofthe transient peak arrivals as t = t + nP + 12 n P ˙ P + 16 n P ˙ P , (1)where t is a reference time, P is the “mean” period, ˙ P is the period derivative, and n is the peak number start-ing from the first peak set as n=0. We use MCMC esti-mates of the 1- σ parameter uncertainties. When fitting without a ˙ P initially, we found the best-fit parameter P = 111.20 ± χ of 9.75 for 14 degrees of freedom. The fit noticeablyimproved when including a ˙ P with t = 3.4 +2 . − . days, P = 114.6 ± P = − ± χ of 2.07. This fit is shown in Figure 2 as thesolid line. If we expand the time of peak uncertaintiesin quadrature by 0.63 days to have a χ per degree offreedom of unity, we find best-fit parameters t = 5.9 ± P = 114.2 ± P = − ± ± ± Payne et al. F l u x [ e r g ss c m ] Figure 3.
The stacked ASAS-SN V -band (green), ASAS-SN g -band (blue) and TESS (maroon) light curves as a function ofphase. The light curves are offset for clarity and the data of each phase are given a different color shade. The right panel showsthe phase-stacked light curves binned at 5 day intervals for ASAS-SN and binned at 1 day intervals for
TESS . The outburstpeaks align across these filters and the shapes are similar. correspond to 2020 September 7.4 ± ± V - and g -band ASAS-SN light curves forASASSN-14ko stacked using the phasing of this modelare shown in Figure 3 along with the TESS light curve.Including the ˙ P significantly reduces the scatter betweenthe stacked light curves. We also binned these stackedlight curves. As apparent in the right panel of Figure3, the peak times are closely aligned across these threefilters and the light curve morphologies are similar be-tween peaks and these filters. The outbursts are char-acterized by a fast rise to peak followed by a shallowerdecline.The Catalina Real-Time Transient Survey (CRTS;Drake et al. 2009) data contain observations spanningnine years prior to the start of the ASAS-SN V -bandlight curve. However, the CRTS data quality made itdifficult to identify prior outburst peaks relative to thequiescent baseline. The CRTS photometry includes fluxfrom the entire host galaxy since it uses Source Extrac-tor (Bertin & Arnouts 1996) rather than image subtrac- tion. Since the host galaxy is bright and spatially large,it is difficult to recognize prior outbursts. We also foundthat data from the All-Sky Automated Survey (ASAS;Pojmanski 1997) were not useful for identifying earlieroutbursts and we could find no other earlier data.4.2. TESS Light Curve Analysis
The
TESS light curve of ASASSN-14ko’s November2018 outburst provides a unique view of the outburstmorphology. Because
TESS observed ASASSN-14koover three consecutive sectors, the light curve capturesthe pre-outburst quiescence, the full rise to peak, andthe full decline back to quiescence. We first character-ized the early-time rise of ASASSN-14ko with a powerlaw model of f = z when t < t , and (2) f = z + h (cid:18) t − t days (cid:19) α when t > t , (3)consisting of residual background z , the time of rise t ,a flux scale h , and the power law index α . We use SASSN-14ko Table 3.
Observed time and maximum light of each out-burst observed by ASAS-SN in the V- and g-band. Shown inblue are the predicted times of the next two outburst peaksin the optical, based on our standard model with a periodderivative.JD Flux at Peak (mJy) O-C (days)2456861.2 +4 . − . +1 . − . − +4 . − . +15 . − . +3 . − . − +15 . − . +3 . − . +1 . − . − +3 . − . +6 . − . +0 . − . − +6 . − . +4 . − . +7 . − . − +4 . − . +3 . − . +1 . − . − +3 . − . +9 . − . +1 . − . − +9 . − . +6 . − . +1 . − . +0.4 +6 . − . +5 . − . +0 . − . − +5 . − . +1 . − . +0 . − . − +1 . − . +0 . − . +0 . − . − +0 . − . +1 . − . +0 . − . +0.5 +1 . − . +0 . − . +0 . − . +0.7 +0 . − . +1 . − . +0 . − . − +1 . − . +0 . − . +0 . − . − +0 . − . +2 . − . +0 . − . − +2 . − . +1 . − . ... ...2459210.0 +1 . − . ... ... the package SCIPY . OPTIMIZE . CURVE FIT’s (Virta-nen et al. 2020) Trust Region Reflective method to ob-tain a best fit model with parameters z = − . ± . − cm − , h = 0 . ± .
01 ergs s − cm − , t =2458429 . ± .
05 JD, and α = 1 . ± .
07. This fitis shown as the red curve in Figure 4. For this fit weinflated the error bars in quadrature by 0.0086 ergs s − cm − in order to make the reduced χ of the fit unity for160 degrees of freedom. The high photometric precisionof TESS shows that the early-time rise was smooth, andthat the time to peak in the
TESS filter is 5 . ± . t when measuring the peak from the databinned at 8-hour intervals.Three TDEs have estimates of α , all of which aresteeper. The first TDE detected by TESS was ASASSN-19bt, which had a power law index of α = 2 . ± . α = 1 . ± . α ∼ α = 2 . ± . TESS light curvealso reveals that the decline was also remarkablysmooth, as shown in Figure 5. Since the rate of the decline in partial TDEs is predicted to be steeper thanthe canonical t − / model, we fit the TESS light curvestarting from five days after peak as f = z − h (cid:18) t − t days (cid:19) α (4)with t being the time of disruption, constrained to bebefore the start of the rise, t , determined above. Weagain inflated the error bars in quadrature by 0.01 ergss − cm − in order to make the reduced χ unity for 2611degrees of freedom. The best-fit power-law has the pa-rameters t = 2458429 . ± .
45 JD, which correspondsto the start of the rise, z = 0 . ± .
001 ergs s − cm − , h = − . ± .
64 ergs s − cm − , and α = − . ± . α = − / − .
66, is actually shallower.We also fit the decline as an exponential decay of theform f = ae − ( t − t peak ) /τ + c (5)which returns the best-fit parameters a = 0 . ± . − cm − , τ = 12 . ± .
11 days, and c = 0 . ± .
001 days. We set t peak to the peak of the TESS lightcurve since its value is degenerate with the other pa-rameters. This model has a reduced χ value of 0.77for the same uncertainties. The exponential decline is,therefore, a better representation of the decline than thepower-law decay model. Since starting the fit five daysafter peak is somewhat arbitrary, Table 4 gives the re-sults for fits starting 5, 10, 15, and 20 days after peakall for the same error model. The fit parameters are rel-atively stable and the exponential model is always thebetter fit. 4.3. May 2020 Outburst
The May 2020 outburst was the first to be predictedin advance, and it occurred as expected. We initiallypredicted the outburst to peak on MJD 58990.2 ± g -band light curve actuallypeaked on MJD 58987.5 +2 . − . , as shown in the last panelof Figure 1, consistent with the prediction.We requested Swift observations to monitor the out-burst along with the LCOGT and amateur ground-baseddata. The non-host-subtracted light curves along withthe X-ray hardness ratios are shown in Figure 6. Unfor-tunately, ground-based observations were severely im-pacted by COVID-19 closures and the impending Sunconstraint. The ASAS-SN light curve in the monthleading up to the May 2020 outburst had a gap dueto these closures. Observations were collected by theamateur astronomers, but large gaps exist in those light2 Payne et al.
JD - 24584000.00.10.2 F l u x [ e r g ss c m ] R e s i d u a l s [ e r g s s c m ] Figure 4.
The rising phase of the
TESS image subtraction light curve. The best-fit power-law model for the rise until JD2458431.5 is shown in red. The bottom panel shows the flux residuals. The
TESS data binned by four hours are shown byturquoise squares. F l u x [ e r g ss c m ]
10 20 30 40 50 60 70 80 90JD - 24584000.0250.0000.025 R e s i d u a l s [ e r g s s c m ] Figure 5.
The declining phase models of the
TESS light curve starting 5 days past peak. The best-fit power-law decline isshown in red and the best-fit exponential decline is shown in blue. The
TESS data binned in 8 hour intervals are shown byturquoise squares. The bottom panel shows the residuals of the fits color-coded by model. Magenta and cyan squares show theresiduals of the power-law decline and exponential decline compared to the binned
TESS data, respectively.
SASSN-14ko Table 4.
Best-fit parameters for the power-law and exponential decline models of the
TESS light curve starting the fits fordifferent numbers of days after peak. z for the power-law decline was 0 . ± .
001 ergs s − cm − for all start times post-peak. Power-Law Decline
Fit Start Timefrom Peak (Days) t (JD-2458400) h (ergs s − cm − ) α χ per dof5 29 . ± . − . ± . − . ± .
03 1.0910 29 . ± . − . ± . − . ± .
05 0.8415 29 . ± . − . ± . − . ± .
11 0.7620 29 . ± . − . ± . − . ± .
20 0.76
Exponential Decline
Fit Start Timefrom Peak (Days) a (ergs s − cm − ) τ (days) c (days) χ per dof5 0 . ± .
001 12 . ± .
11 0 . ± .
001 0.7710 0 . ± .
001 10 . ± .
13 0 . ± .
001 0.7015 0 . ± .
001 11 . ± .
32 0 . ± .
001 0.6520 0 . ± .
001 10 . ± .
49 0 . ± .
001 0.64 Payne et al. curve due to weather closures. The combined data setwas processed as uniformly as possible, but the scatterprevents unambiguous interpretations of the multi-bandlight curve.The most interesting feature of Figure 6 is that weclearly see a wavelength dependence to the flux peak.While the rise was fully observed for the B -band andlonger wavelengths, the U V and U -band data clearlypeaked at still earlier times and the optical bands laggedthe U V -bands by several days. This is also apparentin the blackbody fits to the host-subtracted
Swift datashown in Figure 7 and Table 5. The peak blackbody lu-minosity and temperature occurred approximately fourdays prior to the ASAS-SN g -band peak. However, theblackbody radius remained roughly consistent over time.The Swift XRT X-ray light curve did not follow thesame trend. At peak optical flux, the X-ray flux haddropped by a factor of ∼ g -band peak, which approximatelycorresponds to the U V peak and post-maximum
U V quiescence. The peak luminosity is roughly 1% of theEddington luminosity derived in Section 3.4.4.
Comparison to TDEs
We first compared each of ASASSN-14ko’s out-bursts to the peak luminosity-decline rate relation forpreviously-studied TDEs (Hinkle et al. 2020a). This re-lation describes a correlation between the peak lumi-nosity and its decline luminosity over 40 days. First, webolometrically corrected the ASAS-SN V - and g -band,and TESS photometry using a temperature of 28,800K, which is the median temperature of the most recentoutburst based on the
Swift data. Then, following theprocedure of Hinkle et al. (2020a), we calculated a peakluminosity and the decline rate over 40 days for eachoutburst. In Figure 9, we compare the peak luminos-ity and decline rate of these outbursts to known TDEs.Even though the power-law slope of the decline is shal-lower than t − / (Section 4.2), the actual decline rate issteeper than all TDEs of similar luminosity.We directly compare the TESS light curves ofASASSN-14ko and ASASSN-19bt (Holoien et al. 2019a) in Figure 10. ASASSN-14ko clearly evolves much morerapidly than ASASSN-19bt, with both a more rapid riseand a more rapid decline. However, the overall mor-phologies of the declining light curves are very similarif we compress the time relative to peak of ASASSN-19bt by 30% in order to align the light curve declines asshown in the right panel of Figure 10. ANALYSIS OF SPECTRA5.1.
Evolution of the Optical Spectra
We used optical spectra observed at different points intime and shown in Figure 11 to first classify ASASSN-14ko using standard BPT diagnostics (Baldwin et al.1981; Veilleux & Osterbrock 1987; Kauffmann et al.2003; Kewley et al. 2001; Kewley et al. 2006). The 1996spectrum and a weighted average of the 2014-2015 spec-tra both have line ratios consistent with an AGN. Themeasured ratios are given in Table 6.Next we examined the evolution of the H β and H α emission-line profiles. The spectra from 2014-2015 weretaken after ASASSN-14ko’s optical peak, as indicatedby orange tick marks in Figure 1. There is a noticeablechange in the the Balmer line strengths and profiles dur-ing outburst. Since the line changes associated with thetransient are at the redshift of ESO 253 − G003, we canbe certain that these transients are not Galactic in ori-gin. Directly comparing these spectra should be donewith caution because the data were taken with differentinstruments and long-slit setups. However, it is appar-ent that the emission-line profile shapes clearly changedduring the 2014 November outburst.We obtained five spectra with the LCOGT FLOYDSspectrograph during quiescence in April 2020 and twospectra during the optical rise of the May 2020 out-burst. We compared them to determine any change inthe emission-line profiles as shown in Figure 12. Dueto observatory airmass constraints and ASASSN-14ko’slow position near the horizon from Siding Springs Ob-servatory at that time, observations were not possiblepast 2020-05-17. These two observations occur one andtwo days before the optical peak, respectively, whichcoincides with the start of the decline in the UV, asshown by the vertical dashed lines in Figure 6 and theorange tick marks in Figure 1. A noticeable feature ofthe spectra during outburst is a blue wing around H β near 4830˚A that is not present in the spectra taken dur-ing quiescence. This feature is similar to the broad-ened wings present in the 2014-11-16 and 2014-25 spec-tra which were taken during an outburst optical decline.This gives further evidence that the Balmer lines changeduring the outburst. SASSN-14ko F l u x [ e r g s s c m ] H a r d n e ss R a t i o H S H + S F l u x [ e r g ss c m ] JD - 2456850
Figure 6.
Non-host subtracted photometry of the May 2020 outburst. ASASSN-14ko g -band photometry is represented assquares, Swift data as circles, LCOGT data as pentagons, and data from amateur observatories as diamonds. B -band, g -band,and V -band data taken from different telescopes on the same day were averaged. Swift B - and V -band data were converted toJohnson B and V magnitudes before being converted to flux to enable direct comparison with the ground-based data. Errorbars are plotted but are frequently smaller than the size of the points. The red shaded region denotes the time of the ASAS-SN g -band peak on JD 2458987 . +2 . − . and the magenta vertical solid line shows the predicted peak for our model with a periodderivative, as was also shown in Figure 1. The vertical dashed lines indicate the observation times of the LCOGT spectra shownin Figure 12. Payne et al. l o g [ L ( e r g s s )] l o g [ R ( c m )] l o g [ T ( K )] Figure 7.
Evolution of the UV/optical blackbody lumi-nosity (top panel), radius (middle panel), and temperature(bottom panel) during the May 2020 outburst based on thehost-subtracted
Swift data and shown in Table 5. The timeis relative to the g -band peak on MJD 58987.5. The lumi-nosity and temperature peak occurred several days prior tothe g -band peak. Table 5.
Blackbody luminosity, radius, and tempera-ture during the May 2020 outburst derived from the host-subtracted
Swift data.MJD log [L (ergs s − )] log [R (cm)] log[T (K)]58983 44 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . The similarities between the spectroscopic evolutionof the two outbursts separated by nearly six years indi-cates that the spectroscopic change during the flare maybe a consistent component of these events. The photo-metric and spectroscopic evolution appear closely tied, log[ ( Å )] l o g [ L ( e r g s s )] edd edd Near Peak (MJD = 58983)Post Peak (MJD = 58994)
Figure 8.
Host-subtracted spectral energy distribution attwo epochs during the May 2020 outburst. The first epochon MJD 58983 (blue) is four days prior to the g -band peakon MJD 58987.5. The second epoch on MJD 58994 (red) isseven days after the g -band peak. For the UV/optical, thedata are shown as points while the lines represent the best-fitblackbody components. The X-ray luminosities are shown atthe center of the 0.3-10 keV band over which the luminosityis determined. The dashed gray and dotted gray lines are1% and 10% respectively of the Eddington luminosity for anSMBH of mass 7 . × M (cid:12) . L l o g [ L p e a k / ( e r g s s )] TDEsASAS-SN VASAS-SN gTESS
Figure 9.
The peak luminosity-decline rate for ASASSN-14ko’s outbursts compared to twenty-one previously studiedTDEs. Individual ASAS-SN V (green circles), ASAS-SN g (teal circles), and TESS (red circle) epochs are shown alongwith the TDEs analyzed by Hinkle et al. (2020a) representedas gray squares. The black solid line is the best-fit line forthe TDEs and the dashed black lines are the allowed rangeof uncertainty from the best fit line.
SASSN-14ko
50 0 50 100
Time relative to peak [Days] N o r m a li z e d F l u x
20 0 20 40
Time relative to peak [Days] N o r m a li z e d F l u x Figure 10.
The
TESS light curves of ASASSN-14ko (red circles) binned in 4 hour intervals and ASASSN-19bt (blue diamonds)binned in 16 hour intervals. The left panel shows the light curves as observed, and the right panel shows the ASASSN-19btlight curve compressed in time by 30% to align the light curve declines.
Table 6.
Diagnostic emission line ratios log ([O III]/H β ),log ([N II]/H α ), log ([S II]/H α ), and log ([O I]/H α ) usedto distinguish AGN from H II-regions and classify AGN aseither Seyferts or LINERs (Baldwin et al. 1981, Veilleux &Osterbrock 1987, Kewley et al. 2001, Kauffmann et al. 2003,Kewley et al. 2006). The 2014-2015 ratios were measuredfrom a weighted average of the 2014-2015 spectra.Ratio Diagnostic 1996 2014-2015log ([O III]/H β ) 0.73 ± ± ([N II]/H α ) − ± − ± ([S II]/H α ) − ± − ± ([O I]/H α ) − ± − ± and higher temporal resolution of upcoming outburstsmay reveal further connections between the rise and de-cline of the photometric light curves with morphologicalchanges in the Balmer lines.5.2. X-ray Analysis
We compared the two X-ray spectral epochs to charac-terize the X-ray emission evolution and are shown in Fig-ure 13. We modeled the 2015
XMM-Newton + NuSTAR spectra as a combination of a soft black body plus apower law. The black body model had a temperatureof 0 . ± .
01 keV and the power law model had aΓ = 0 . ± .
04. There is a strong 6.4 keV Fe linepresent that we include in the model as a Gaussian. The
Swift
XRT spectrum was extracted from the mergeddata taken from ∼
30 days prior to ∼ . +0 . − . and the fit was notimproved by adding an additional black body with a best-fit temperature of 0 . ± .
03 keV. The
Swift spec-trum had combined all the May 2020 observations be-cause the signal-to-noise ratio of the spectrum wouldotherwise have been too low to assess its characteristics.With only ∼
650 counts, poor statistics likely preventany detection of the Fe line. The fit parameters aresummarized in Table 7.The
XMM-Newton spectrum was taken during a pe-riod when the optical flare was quiescent, however theflux derived from the
XMM-Newton spectrum is similarto what we observed with
Swift . The similarities be-tween the
Swift spectrum and the
XMM-Newton spec-trum may indicate that even during times around theUV peak, the average properties of the X-ray emissionduring May 2020 may not have changed significantlyfrom the properties of the deep
XMM-Newton observa-tion. Future observations with better signal-to-noise willbe essential to disentangle potential X-ray emission evo-lution between quiescence and outburst. There are nineepochs of
XMM-Newton slew observations of this galaxyobserved between 2007-2019. However, given the sparsesampling and low signal to noise, there is no obviouspattern after phase folding these data. DISCUSSIONHere we discuss several scenarios for ASASSN-14ko’speriodic outbursts and some of their problems. We con-sider the possibilities that ASASSN-14ko is a sub-parsecSMBH binary system, a SMBH and perturbing mas-sive star binary system, and, finally, a repeating partialTDE. We do not consider stellar origins for the out-bursts. As noted earlier, a Galactic source is ruled outbecause the emission line changes associated with theoutbursts occur at the redshift of the host. At theredshift of the host, the peak luminosities are similar8
Payne et al. F l u x [ e r g ss c m Å ] F l u x [ e r g ss c m Å ] H- F l u x [ e r g ss c m Å ] H- Figure 11.
Spectra of ESO 253 − G003 showing the change in the Balmer emission-line profiles. The earliest epoch from 1996-02-19 (Kewley et al. 2001) is shown in violet. Subsequent spectra were taken after the discovery of ASASSN-14ko including thespectrum reported in Holoien et al. (2014b) and five epochs from PESSTO. The spectra have been scaled using mapspec to putall of them onto a common flux scale. F l u x [ e r g ss c m Å ] F l u x [ e r g ss c m Å ] H-H-H-H-H-H-H- F l u x [ e r g ss c m Å ] H-H-H-H-H-H-H-
Figure 12.
Spectra of ESO 253 − G003 observed with the FLOYDS spectrograph at LCOGT during quiescence in April 2020and during the May 2020 outburst. The first five spectra were obtained during quiescence and the last two spectra were obtainedduring the optical rise of May 2020 outburst. The spectra have been scaled using mapspec to put all of them onto a commonflux scale.
SASSN-14ko P h o t o n s [ c m s k e V ] Energy (keV) R a t i o P h o t o n s [ c m s k e V ] Energy (keV) R a t i o Figure 13.
XMM-Newton + NuSTAR spectra (left) from MJD=57253.6 during a period of quiescence compared to the recentSwift spectra (right). The recent Swift observations do not show the Fe line around 6.4 keV, but this is likely due to the lowsignal-to-noise ratio. to those of luminous supernovae and so are too highfor non-explosive stellar transients. The many quasi-periodic repetitions of similar luminosity rule out explo-sive possibilities. This appears to leave only phenomenaassociated with a central SMBH as possible explana-tions.6.1.
ASASSN-14ko as a SMBH Binary System
As discussed in the introduction, apparently periodicemission from AGN is frequently interpreted as evi-dence for an SMBH binary. However, there are threemain inconsistencies with this scenario for ASASSN-14ko: (1) the period derivative; (2) the short life timeof such a binary and (3) the lack of velocity shifts. Wewill scale the results for black hole masses of M BHa =5 × ˆ M BHa M (cid:12) , M BHb = 5 × ˆ M BHb M (cid:12) and a pe-riod of P = 114 ˆ P days.Due to the emission of gravitational waves, the binarywill have a period derivative of˙ P = 0 . f ( e ) ˆ M BHa ˆ M BHb ˆ P / (cid:16) ˆ M BHa + ˆ M BHb (cid:17) / , (6)where f ( e ) = (1 + 73 e /
24 + 37 e / − e ) − / is thedependence on the orbital ellipticity e , and a time tomerger of t m = 2100 g ( e ) ˆ P / (cid:16) ˆ M BHa + ˆ M BHb (cid:17) / ˆ M BHa ˆ M BHb years(7)where g ( e ) = (1 − e ) / (Peters 1964). The first prob-lem is that the observed period derivative is over anorder of magnitude larger than the scaling in Eqn. 6 for black holes in a circular orbit. This can be solvedonly by substantially increasing the mass of at least oneof the black holes, or by making the orbit significantlyelliptical. Both of these solutions to the mismatch in pe-riod derivatives will exacerbate the next two problemsby reducing the binary lifetime and increasing the bi-nary velocities. If the solution is to change the masses,the lines must be formed around the less massive SMBH,which also exacerbates the velocity problem.The second problem is that finding a binary so close tomerging is very improbable. The first way to phrase thisis through what it requires for the properties of SMBHbinaries in other galaxies. The number density of L ∗ galaxies is roughly n (cid:39) . h Mpc − (e.g., Kochaneket al. 2001), so there are roughly N g (cid:39) × suchgalaxies inside the distance to ESO 253 − G003. To haveone system merging in the next t = 1000 years im-plies that all of these galaxies must contain binaries thatwill merge in the next N g t (cid:39)
300 million years. Thisin turn implies that they must all have orbital periodsshorter than ∼ N / g P (cid:39)
35 years, which would seemto make the problem of finding binary SMBH systemsrather trivial rather than being as difficult as it appearsto be in practice. The second way to phrase the prob-lem is that if there is one SMBH binary with a period < / / = 19 with pe-riods < < <
10 years.The third problem for an SMBH binary is the highvelocity scale, with v a (cid:39) (cid:34) ˆ M BHa + ˆ M BHb ˆ P (cid:35) / ˆ M BHb sin i ˆ M BHa + ˆ M BHb km/s(8)0
Payne et al.
Table 7.
Best-fit parameters for the
Swift and
XMM-Newton + NuSTAR
X-ray spectra. Both models used a redshift of 0.0425and a neutral hydrogen column density N H of 3 . × cm − frozen to the Galactic column density (HI4PI Collaborationet al. 2016). The XMM-Newton + NuSTAR spectrum included a Gaussian line energy of 6 . ± .
03 keV to capture the 6.4 keVFe line. The
Swift fit had 56 degrees of freedom and the
XMM-Newton + NuSTAR fit had 470 degrees of freedom.Instrument kT (keV) BlackbodyNormalization (K) Photon Index Γ χ per dof Swift . ± .
03 79 . +137 . − . . +0 . − . . XMM+NuSTAR . ± .
01 83 . +32 . − . . ± .
04 1.64 for circular orbits. While our phase sampling is poor,we arguably can set a limit on any H β line shifts of < < P is correct, this prob-lem cannot be solved by reducing the masses or invokinga large mass ratio with the emission lines being formedby material associated with the more massive SMBH.Similarly, raising the ellipticity to solve the ˙ P problemmakes this problem worse because of the higher veloc-ities at pericenter compared to a circular orbit of thesame period. The velocity problem can only be solvedby making the system nearly face on or by relying onthe poor spectral sampling to hide the velocity shifts.6.2. ASASSN-14ko as a SMBH with a Perturbing Star
Rather than a binary system with two SMBHs, theperiodic outbursts could be driven by a star orbitinga single SMBH. Because stars are far more commonthan SMBHs and the gravitational wave merger lifetimesare now far longer, the probability argument against anSMBH binary is removed. The long gravitational wavemerger time does mean that the observed ˙ P must havea different origin such as viscous interactions betweenthe disk and the star, although estimating this effect isnon-trivial.We now assume a star of mass M ∗ = ˆ M ∗ M (cid:12) and ra-dius R ∗ = ˆ R ∗ R (cid:12) orbiting a black hole of mass M BH =5 × ˆ M BH M (cid:12) . There are three length scales of im-mediate interest, and we show their relative values andtheir dependence on SMBH mass in Figure 14. TheSchwarzschild radius of the black hole is R s = 1 . × ˆ M BH cm , (9)the tidal (Roche) radius to disrupt the star is R T = 2 . × ˆ R ∗ ˆ M − / ∗ ˆ M / BH cm (10)and the orbital semi-major axis is a = 2 . × ˆ P / ˆ M / BH cm . (11)Finally, for a pericentric radius R p , the ellipticity of theorbit is e (cid:39) − . R p R T ˆ R ∗ ˆ M / ∗ . (12)
12 13 14 15 1677.27.47.67.88
Figure 14.
The Schwarzschild radius ( R s , solid), semima-jor axis ( a , dotted) and several tidal radii ( R T , dashed)as a function of SMBH mass M BH . The tidal radii arefor a 0 . M (cid:12) main-sequence star ( R T (0 . M (cid:12) )), the Sun( R T (1 . M (cid:12) )), and a star with the mass of the Sun and a ra-dius of 10 R (cid:12) ( R T ( M (cid:12) , R (cid:12) ). On these logarithmic scales,the effects of spin on the BH horizon and the tidal limits aremodest. For some star/SMBH scenarios, the true period wouldbe twice the observed period, which would increase thesemimajor axis by 0 . ωP π = 6 πGM BH ac (1 − e ) (cid:39) R s R p , (13)for a Schwarzschild black hole. This means that wewould expect systematic changes with time independentof the mechanism driving the flares. SASSN-14ko
20 0 20 40 60 80 100Time relative to peak [Days]1.01.52.02.53.0 F l u x [ e r g ss c m ] Figure 15.
The stacked ASAS-SN g -band light curves separated by even (blue) and odd (red) outbursts as a function of phase.The light curves are offset for clarity and the data of each phase are given a different color shade. The binned even light curve,shown by star symbols, is superimposed over the odd light curves, and vice versa, with the binned odd light curve shown bydiamond symbols. If the star is not being tidally disrupted, then the flaresmust be driven by periodic disturbances of the accretionflows as the star passes through the accretion disk. How-ever, a star simply embedded in the disk would repre-sent a continuous perturbation that is unlikely to driveperiodic flares. Flaring would seem to require a stel-lar orbit at a significant inclination angle relative to thedisk. The star would then make two passages throughthe disk per orbit, so the orbital period would likely betwice the flare period.Each orbit would produce a pair of flares with thespacing dependant on the orbital eccentricity and theargument of periapsis ( ω ) relative to the accretion disk.The separation between pairs will increase as the orbitbecomes more eccentric. While eccentric orbits with ω near 0 or π can have equally spaced encounters, the starwill encounter the disk at different radii/temperatures.Given that the flare spacing, profiles, and amplitudesare all essentially constant between cycles, a perturbingstar would seem be required to be on an inclined butnearly circular orbit. Even then, the similarity of theflares seems odd because one would expect encounterswith the star moving away from the observer and into the disk to differ from the reverse, except for nearlyedge on viewing angles. In Figure 15 we see that theeven and odd flare profiles are very similar in amplitude,shape, and duration which is difficult to reconcile withthis model.Finally, while an inclined, circular orbit of a star orbit-ing a SMBH might be able to perturb the accretion diskwith the right frequency, there is no obvious time scalein disks to then make the flares so short in duration.6.3. ASASSN-14ko as a Repeated Partial TDE
A third possibility to explain ASASSN-14ko’s peri-odic outbursts is as a repeating TDE that is partiallydisrupted after each passage close to the central SMBH.While it requires fine tuning to have a main sequencestar pass close to its tidal limit but remain outside theSMBH horizon, we can see from Figure 14 that it is rel-atively easy for an evolved star on an elliptical orbit todo so. As discussed in the introduction, giants are alsothe most likely candidates for partial disruptions thatcould power periodic flares. MacLeod et al. (2012) andGuillochon & Ramirez-Ruiz (2013) find that the starwill begin to lose mass once β = R T /R p < . Payne et al.
The peak luminosities of the flares, L p (cid:39) × erg/s correspond to peak accretion rates of ˙ M p (cid:39) . (cid:15) − . M (cid:12) /year where (cid:15) = 0 . (cid:15) . is the accretion effi-ciency. If the peaks last ∼
10 days, the accreted massof ∆ M (cid:39) . (cid:15) − . M (cid:12) is certainly low enough to allowrepeated outbursts on this scale. Note, however, thatthese accretion rates are significantly higher than envi-sioned by MacLeod et al. (2013) and the time scales aremuch shorter.Ryu et al. (2020) found that the change in the or-bital specific energy of the star in a partial disruption is f ∼ − of the specific energy scale GM BH R ∗ /R t ofthe stripped debris. They find both positive and nega-tive energy changes, so there is no prediction of the signof the changes. The period derivative measured in Sec-tion 4.1 implies a change in the orbital specific energyof GM BH ˙ P / a . This means that we should expect aperiod derivative of | ˙ P | = 3 f aR ∗ R t (cid:39) . f ˆ M / ∗ ˆ P / ˆ M / BH ˆ R ∗ . (14)For f ∼ − , this implies | ˙ P | ∼ − with relativelylittle sensitivity to the exact values of the parametersand remarkably close to the measured period derivative.The agreement is perhaps more remarkable because theRyu et al. (2020) simulations were for a single pericentricpassage of a main sequence star on an initially parabolicorbit, rather different from the orbit required here. How-ever, the orbit of the puffy stellar merger remnant in Fig-ure 13 (top) of Antonini et al. (2011) has a semi-majoraxis shrinking as ∆ a/a ∼ − per orbit, which is thesame order of magnitude. The example in the lowerpanel of this figure shows very little orbital evolutionbut also shows very little ongoing mass loss. Note thatthe orbital changes essentially occur with the pericenterfixed because the tidal interactions are only importantat pericenter. Because the structure of the star mustbe changing with each pericentric passage due to themass loss, torques and heating, ˙ P presumably cannotbe constant on longer time scales.The partial TDE hypothesis also seems better ableto explain the similarity of the flares since each peri-centric passage is almost identical in geometry to theprevious and the required mass loss rates appear to bemodest. However, they cannot truly be identical sincethe orbital geometry must slowly change due to preces-sion (relativistic and tidal) and the mass lost over tensof encounters ceases to be modest. Overall, the repeat-ing, partial TDE interpretation seems most consistentwith the available observations. SUMMARYAlthough ASASSN-14ko was first thought to be a su-pernova, the subsequent six years of ASAS-SN V - and g -band data show that the flares occur at regular intervals.The 17 flares observed to date are well modeled using aperiod of P = 114 . ± . P = − . ± . . ± . . ± . Swift multi-wavelength photometric data,
TESS observed ASASSN-14ko during its November 2018 outburst. The
TESS light curve has a decline rate that is dissimilar to pre-viously studied TDEs and a rapid rise to peak occur-ring over 5 . ± .
05 days. The
TESS data also showthat the rise and decline were smooth and lack short-timescale variability. The individual outbursts are mor-phologically very similar over the six year baseline ofobservations. While a host of problems interfered withstudying the May 2020 outburst well, there was clearevidence that the outburst peaks a few days earlier inthe UV than in the optical. Spectra taken during andprior to the May 2020 outburst revealed morphologicalchanges around H β during the flare which was similarto what occurred during the 2014 outburst. This sug-gests that morphological changes in the emission linesare consistently associated with the optical outburstsover time. We examined several possible scenarios toexplain the cause of this AGN’s unusual behavior, in-cluding the presence of a SMBH binary, a SMBH plusa perturbing massive star, or a repeating partial TDE.Between these scenarios, we favor a repeating partialTDE. We believe that any stellar transients or explo-sions whether Galactic or in the host are ruled out. Themost important next step is to time and study the flaresmore closely across the electromagnetic spectrum. Therelatively short period and system brightness make thisrelatively easy. ASASSN-14ko will be observed by TESS again in Sectors 31-33 during the predicted December2020 outburst. These observations will give further con-straints on the nature of these outbursts, and presents aunique opportunity to do a detailed reverberation map-ping analysis of the system.
Software: ftools (Blackburn 1995),
NuSTAR
DataAnalysis Software (v1.8.0)
SASSN-14ko
Aguero, E. L., Paolantonio, S., & Suarez, F. 1996, PASP,108, 1117, doi: 10.1086/133844Alard, C. 2000, A&AS, 144, 363, doi: 10.1051/aas:2000214Alard, C., & Lupton, R. H. 1998, ApJ, 503, 325,doi: 10.1086/305984Antonini, F., Lombardi, James C., J., & Merritt, D. 2011,ApJ, 731, 128, doi: 10.1088/0004-637X/731/2/128Arcavi, I., Gal-Yam, A., Sullivan, M., et al. 2014, ApJ, 793,38, doi: 10.1088/0004-637X/793/1/38Auchettl, K., Ramirez-Ruiz, E., & Guillochon, J. 2018,ApJ, 852, 37, doi: 10.3847/1538-4357/aa9b7cBade, N., Komossa, S., & Dahlem, M. 1996, A&A, 309, L35Baldwin, J. A., Phillips, M. M., & Terlevich, R. 1981,PASP, 93, 5, doi: 10.1086/130766 Baumgartner, W. H., Tueller, J., Markwardt, C. B., et al.2013, ApJS, 207, 19, doi: 10.1088/0067-0049/207/2/19Bentz, M. C., & Katz, S. 2015, PASP, 127, 67,doi: 10.1086/679601Bertin, E., & Arnouts, S. 1996, A&AS, 117, 393,doi: 10.1051/aas:1996164Blackburn, J. K. 1995, Astronomical Society of the PacificConference Series, Vol. 77, FTOOLS: A FITS DataProcessing and Analysis Software Package, ed. R. A.Shaw, H. E. Payne, & J. J. E. Hayes, 367Blandford, R., Meier, D., & Readhead, A. 2019, ARA&A,57, 467, doi: 10.1146/annurev-astro-081817-051948Bloom, J. S., Giannios, D., Metzger, B. D., et al. 2011,Science, 333, 203, doi: 10.1126/science.1207150 Payne et al.
Bon, E., Jovanovi´c, P., Marziani, P., et al. 2012, ApJ, 759,118, doi: 10.1088/0004-637X/759/2/118Breeveld, A. A., Curran, P. A., Hoversten, E. A., et al.2010, MNRAS, 406, 1687,doi: 10.1111/j.1365-2966.2010.16832.xBrown, T. M., Baliber, N., Bianco, F. B., et al. 2013,PASP, 125, 1031, doi: 10.1086/673168Burrows, D. N., Hill, J. E., Nousek, J. A., et al. 2005,SSRv, 120, 165, doi: 10.1007/s11214-005-5097-2Burrows, D. N., Kennea, J. A., Ghisellini, G., et al. 2011,Nature, 476, 421, doi: 10.1038/nature10374Burtscher, L., Orban de Xivry, G., Davies, R. I., et al.2015, A&A, 578, A47, doi: 10.1051/0004-6361/201525817Calistro Rivera, G., Lusso, E., Hennawi, J. F., & Hogg,D. W. 2016, ApJ, 833, 98,doi: 10.3847/1538-4357/833/1/98Campana, S., Mainetti, D., Colpi, M., et al. 2015, A&A,581, A17, doi: 10.1051/0004-6361/201525965Cenko, S. B., Bloom, J. S., Kulkarni, S. R., et al. 2012a,MNRAS, 420, 2684,doi: 10.1111/j.1365-2966.2011.20240.xCenko, S. B., Krimm, H. A., Horesh, A., et al. 2012b, ApJ,753, 77, doi: 10.1088/0004-637X/753/1/77Charisi, M., Bartos, I., Haiman, Z., et al. 2016, MNRAS,463, 2145, doi: 10.1093/mnras/stw1838Coughlin, E. R., & Nixon, C. J. 2019, ApJL, 883, L17,doi: 10.3847/2041-8213/ab412dDonley, J. L., Brandt, W. N., Eracleous, M., & Boller, T.2002, AJ, 124, 1308, doi: 10.1086/342280D’Orazio, D. J., Haiman, Z., & MacFadyen, A. 2013,MNRAS, 436, 2997, doi: 10.1093/mnras/stt1787D’Orazio, D. J., Haiman, Z., & Schiminovich, D. 2015,Nature, 525, 351, doi: 10.1038/nature15262Drake, A. J., Djorgovski, S. G., Mahabal, A., et al. 2009,ApJ, 696, 870, doi: 10.1088/0004-637X/696/1/870Evans, C. R., & Kochanek, C. S. 1989, ApJL, 346, L13,doi: 10.1086/185567Farris, B. D., Duffell, P., MacFadyen, A. I., & Haiman, Z.2014, ApJ, 783, 134, doi: 10.1088/0004-637X/783/2/134Fausnaugh, M. M., Denney, K. D., Barth, A. J., et al. 2016,ApJ, 821, 56, doi: 10.3847/0004-637X/821/1/56Gehrels, N., Chincarini, G., Giommi, P., et al. 2004, ApJ,611, 1005, doi: 10.1086/422091Gezari, S., Halpern, J. P., Komossa, S., Grupe, D., &Leighly, K. M. 2003, ApJ, 592, 42, doi: 10.1086/375553Gezari, S., Martin, D. C., Milliard, B., et al. 2006, ApJ,653, L25, doi: 10.1086/509918Gezari, S., Basa, S., Martin, D. C., et al. 2008, ApJ, 676,944, doi: 10.1086/529008 Gezari, S., Heckman, T., Cenko, S. B., et al. 2009, ApJ,698, 1367, doi: 10.1088/0004-637X/698/2/1367Gezari, S., Chornock, R., Rest, A., et al. 2012, Nature, 485,217, doi: 10.1038/nature10990Gold, R., Paschalidis, V., Etienne, Z. B., Shapiro, S. L., &Pfeiffer, H. P. 2014, PhRvD, 89, 064060,doi: 10.1103/PhysRevD.89.064060Graham, M. J., Djorgovski, S. G., Stern, D., et al. 2015a,MNRAS, 453, 1562, doi: 10.1093/mnras/stv1726—. 2015b, Nature, 518, 74, doi: 10.1038/nature14143Greiner, J., Schwarz, R., Zharikov, S., & Orio, M. 2000,A&A, 362, L25. https://arxiv.org/abs/astro-ph/0009430Grupe, D., Komossa, S., & Saxton, R. 2015, ApJL, 803,L28, doi: 10.1088/2041-8205/803/2/L28Grupe, D., Thomas, H. C., & Leighly, K. M. 1999, A&A,350, L31. https://arxiv.org/abs/astro-ph/9909101Guillochon, J., & Ramirez-Ruiz, E. 2013, ApJ, 767, 25,doi: 10.1088/0004-637X/767/1/25Guo, D., Tao, J., & Qian, B. 2006, PASJ, 58, 503,doi: 10.1093/pasj/58.3.503HEASARC. 2014, HEAsoft: Unified Release of FTOOLSand XANADU. http://ascl.net/1408.004Heckman, T. M., & Best, P. N. 2014, ARA&A, 52, 589,doi: 10.1146/annurev-astro-081913-035722Henden, A. A., Levine, S., Terrell, D., & Welch, D. L. 2015,in American Astronomical Society Meeting Abstracts,Vol. 225, American Astronomical Society MeetingAbstracts, 336.16HI4PI Collaboration, Ben Bekhti, N., Fl¨oer, L., et al. 2016,A&A, 594, A116, doi: 10.1051/0004-6361/201629178Hickox, R. C., & Alexander, D. M. 2018, ARA&A, 56, 625,doi: 10.1146/annurev-astro-081817-051803Hills, J. G. 1975, Nature, 254, 295, doi: 10.1038/254295a0Hinkle, J. T., Holoien, T. W. S., Shappee, B. J., et al.2020a, ApJL, 894, L10, doi: 10.3847/2041-8213/ab89a2Hinkle, J. T., Holoien, T. W. S., Auchettl, K., et al. 2020b,arXiv e-prints, arXiv:2006.06690.https://arxiv.org/abs/2006.06690Ho, L. C. 2008, ARA&A, 46, 475,doi: 10.1146/annurev.astro.45.051806.110546Holoien, T. W. S., Brown, J. S., Auchettl, K., et al. 2018,MNRAS, 480, 5689, doi: 10.1093/mnras/sty2273Holoien, T. W.-S., Prieto, J. L., Bersier, D., et al. 2014a,MNRAS, 445, 3263, doi: 10.1093/mnras/stu1922Holoien, T. W. S., Kiyota, S., Brimacombe, J., et al. 2014b,The Astronomer’s Telegram, 6732, 1Holoien, T. W.-S., Kochanek, C. S., Prieto, J. L., et al.2016a, ArXiv e-prints. https://arxiv.org/abs/1602.01088—. 2016b, MNRAS, 455, 2918, doi: 10.1093/mnras/stv2486