Evidence for rapid disk formation and reprocessing in the X-ray bright tidal disruption event AT 2018fyk
We present optical spectroscopic and Swift UVOT/XRT observations of the X-ray and UV/optical bright tidal disruption event (TDE) AT 2018fyk/ASASSN–18ul discovered by ASAS–SN. The Swift lightcurve is atypical for a TDE, entering a plateau after 40 days of decline from peak. After 80 days the UV/optical lightcurve breaks again to decline further, while the X-ray emission becomes brighter and harder. In addition to broad H, He and potentially O/Fe lines, narrow emission lines emerge in the optical spectra during the plateau phase. We identify both high ionisation (O iii) and low ionisation (Fe ii) lines, which are visible for 45 days. We similarly identify Fe ii lines in optical spectra of ASASSN–15oi 330 d after discovery, indicating that a class of Fe-rich TDEs exists. The spectral similarity between AT 2018fyk, narrow-line Seyfert 1 galaxies and some extreme coronal line emitters suggests that TDEs are capable of creating similar physical conditions in the nuclei of galaxies. The Fe ii lines can be associated with the formation of a compact accretion disk, as the emergence of low ionisation emission lines requires optically thick, high density gas. Taken together with the plateau in X-ray and UV/optical luminosity this indicates that emission from the central source is efficiently reprocessed into UV/optical wavelengths. Such a two-component lightcurve is very similar to that seen in the TDE candidate ASASSN–15lh, and is a natural consequence of a highly relativistic orbital pericenter.
keywords:accretion, accretion disks – galaxies: nuclei – black hole physics – ultraviolet: galaxies – X-rays: galaxies
Passing within the tidal radius of the supermassive black hole (SMBH) in the centre of a galaxy can lead to a star’s demise (Hills, 1975; Rees, 1988; Phinney, 1989). Such cataclysmic events, called tidal disruption events (TDEs), resemble panchromatic cosmic fireworks, with bright emission at wavelengths ranging from radio (van Velzen et al., 2016a; Alexander et al., 2016), IR (van Velzen et al., 2016b; Jiang et al., 2016; Mattila et al., 2018), optical and UV (Gezari et al., 2008; van Velzen et al., 2011; Arcavi et al., 2014; Holoien et al., 2016b; Wyrzykowski et al., 2017) as well as X-rays (Komossa & Bade, 1999; Greiner et al., 2000) and even rays (Bloom et al., 2011; Cenko et al., 2012). The duration and brightness of such flares depends on the complex dynamics of material in the presence of strong gravitational fields (Guillochon & Ramirez-Ruiz, 2015; Metzger & Stone, 2016). Wide-field surveys such as the Roentgen Satellite (ROSAT) and the X-ray Multi-Mirror telescope (XMM; Jansen et al. 2001) in X-rays and the Galaxy and Evolution Explorer (GALEX), Sloan Digital Sky Survey (SDSS; Stoughton et al. 2002), the (intermediate) Palomar Transient Factory (PTF; Law et al. 2009) and the All Sky Automated Supernova (ASASSN; Kochanek et al. 2017) surveys in the UV/optical have led to the discovery and characterisation, first in archival data and later in near real-time, of a few dozen TDEs and even more TDE candidates.
Sparse (or non-existent) temporal data coverage of UV/optical selected TDEs at X-ray wavelengths (and vice-versa) inhibit the multi-wavelength characterisation and subsequently the detailed study of the energetics and dynamics at play. This sparse coverage is the result of a variety of factors, such as the difficulty to perform image subtraction in galactic nuclei, the need for fast and systematic spectroscopic follow-up of nuclear transients and the limited availability of multi-wavelength monitoring. Coordinated efforts in recent years have led to significant progress in this respect, and most spectroscopically confirmed TDEs are now observed with the Swift X-ray observatory, made possible due to its flexible scheduling system.
Nevertheless, disentangling the dominant emission mechanisms remains a challenge. The thermal soft X-ray emission is thought to originate from a compact accretion disk (e.g. Komossa & Bade 1999; Auchettl et al. 2017) while luminous hard X-ray emission finds it origin in a relativistic jet (Bloom et al., 2011; Cenko et al., 2012). For the UV/optical emission, however, a clear picture has not yet emerged. Shocks due to stream-stream collisions (Piran et al., 2015; Shiokawa et al., 2015) or reprocessing of accretion power in either static (Loeb & Ulmer, 1997; Guillochon et al., 2014; Roth et al., 2016) or outflowing material (e.g. Strubbe & Quataert 2009; Metzger & Stone 2016; Roth & Kasen 2018) have all been proposed to explain the observations. Dai et al. (2018) proposed a model that can explain both the X-ray and UV/optical observations by suggesting a geometry similar to the active galactic nucleus (AGN) unification model (see also Metzger & Stone 2016), where an optically thick structure in the disk orbital plane or an optically thick super-Eddington disk wind obscures the X-ray emission for certain viewing angles. The presence of Bowen fluorescence lines, which require an X-ray powering source, in several TDEs with X-ray non-detections (Leloudas et al., 2019), support this scenario.
In terms of their optical spectra, TDEs typically show broad (10–20 km s) H and/or He lines (Arcavi et al., 2014), although it is unclear what determines whether a TDE is H-rich, He-rich or shows both features. Furthermore, while some TDEs show only broad He ii emission, the sudden appearance or disappearance of other lines such as He i has been observed (Holoien et al., 2016a). One feature in particular is observed in many TDEs: the broad He ii line appears to have an asymmetric shoulder in its blue wing. Moreover, it is often observed to be significantly blueshifted (when fit with a Gaussian line profile), whereas other broad Balmer lines, when present, do not show a similar blueshift. While asymmetric Balmer emission line profiles can be modelled using an elliptical accretion disk model (Liu et al., 2017; Cao et al., 2018; Holoien et al., 2018) or alternatively a spherically expanding medium (Hung et al., 2019), it does not appear to adequately explain the He ii line morphology. Leloudas et al. (2019) suggest instead that the asymmetry in the line is due to Bowen fluorescence lines, but this cannot explain all cases (e.g. ASASSN–15oi).
Leloudas et al. (2016) were the first to claim that two emission mechanisms were observed in a TDE candidate, namely in the double-peaked lightcurve of ASASSN–15lh. Although the debate as to the nature of this peculiar transient event is still ongoing (Godoy-Rivera et al., 2017; Margutti et al., 2017), one explanation focused on the TDE interpretation. Leloudas et al. (2016) claim that the double-peaked lightcurve can be explained in terms of the fallback and viscous timescales around a very massive (10 M) SMBH. In this case the orbital pericenter of the disrupted star is relativistic, making disk formation very efficient. This can lead to two distinct maxima in the lightcurve. In fact, van Velzen et al. (2018) recently demonstrated that a two-phase structure appears to be common for all TDEs, but often the second, more shallow phase is observed a few years after peak. Alternatively, Margutti et al. (2017) invoke a model where a sudden change in the ejecta opacity due to an underlying source of ionising radiation leads to a double-humped lightcurve. We will show that the lightcurve of AT 2018fyk shows a similarly double-humped profile to ASASSN–15lh. We propose that a very deeply plunging orbit around a less massive black hole can also lead to a relativistic pericenter, speeding up the disk formation process and producing a similar lightcurve.
In this work we present our observations of a new tidal disruption event, AT 2018fyk/ASASSN–18ul, discovered by the All Sky Automated Survey for SuperNovae (ASAS–SN; Shappee et al. (2014)). We analyse Swift UVOT and XRT data together with optical low resolution spectroscopic observations covering the first 120 days of its evolution. While both the lightcurve and spectra show features peculiar to known TDEs, in particular a secondary maximum in the UVOT bands and the simultaneous emergence of narrow emission lines (in addition to broad H and He lines), we show that these properties can be explained by the reprocessing of (part of the) X-ray emission into UV/optical photons. While the lightcurve is similar to ASASSN–15lh, this is the first time that unambiguous evidence for reprocessing is found in the optical spectra of TDEs. This shows that the dynamics of the disruption can leave clear imprints on the lightcurves. Moreover, spectral signatures of reprocessing are only present during the second maximum in the lightcurve. This suggests that the X-ray source turned on after the initial UV/optical peak, in line with a delayed accretion disk formation scenario.
2 Observations and data reduction
The transient AT 2018fyk/ASASSN–18ul was discovered near the center of the galaxy LCRS B224721.6-450748 (estimated offset of 0.85 arcsec from the nucleus) by the ASAS–SN survey on 2018 September 8 (MJD 58 369.23). The estimated transient brightness was =17.8 mag, with a non-detection reported (17.4 mag) on 2018 August 29. A classification spectrum was taken as part of the extended Public ESO Spectroscopic Survey for Transient Objects (ePESSTO; Smartt et al. 2015) on 2018 September 15, revealing a blue featureless continuum superposed with several broad emission lines, suggesting that the transient was likely a TDE (Wevers et al., 2018).
No high spatial resolution archival imaging is available to constrain the position of the transient with respect to the host galaxy centre of light. Fortunately, Gaia Science Alerts (GSA; Hodgkin et al. 2013) also detected the transient (aka Gaia18cyc) at the position (,) = (22:50:16.1, –44:51:53.5) on 2018 October 10, with an estimated astrometric accuracy of 100 mas111This is due to the fact that GSA uses the initial data treatment astrometric solution (Fabricius et al., 2016). In the future, the implementation of an improved astrometric solution could improve this to mas precision.. The host galaxy is part of the Gaia Data Release 2 (GDR2) catalogue (Gaia Collaboration et al., 2016, 2018), and its position is reported as (,) = (22:50:16.093, –44:51:53.499) with formal uncertainties of 1.1 and 1.5 mas in right ascension and declination, respectively (Lindegren et al., 2018). We note that the GDR2 astrometricexcessnoise parameter is 11 mas, which indicates that the formal errors are likely underestimated (as expected for an extended source, Lindegren et al. 2018). The offset between the transient and host galaxy positions is 15 mas.
Kostrzewa-Rutkowska et al. (2018) have shown that the mean offset in the Gaia data of SDSS galaxies is 100 mas, consistent with the mean offset of SDSS galaxies and their GDR2 counterparts. Additionally, we can try to estimate a potential systematic offset between Gaia transients and their GDR2 counterparts. To quantify such an offset, we crossmatch the 7000 published Gaia alerts with GDR2 within a search radius of 0.25 arcsec. The offset distribution (angular distance on the sky) is well described by a Rayleigh function, as expected if the uncertainties in right ascension and declination follow a normal distribution. The distance distribution has a median of 62 mas and standard deviation of 40 mas. This represents the potential systematic offset between the coordinate systems and is fully consistent with the 100 mas transient positional uncertainties, indicating that both coordinate systems are properly aligned.
In conclusion, we find an offset between the transient and host galaxy position of 15100 mas, which corresponds to 17120 pc at the host redshift. This illustrates the power of Gaia for identifying nuclear transients (see also Kostrzewa-Rutkowska et al. 2018 for a detailed investigation), as it firmly constrains AT 2018fyk to the nucleus of the galaxy.
2.1 Host galaxy spectral energy distribution
We determine the host galaxy redshift from the spectra, which show strong Ca ii H+K absorption lines, and find z=0.059. This corresponds to a luminosity distance of approximately 275.1 Mpc, assuming a CDM cosmology with , and (Planck Collaboration et al., 2014). No narrow emission lines from the host galaxy are evident, indicating that the event occurred in a quiescent galaxy. We observe H and H in absorption, indicating no ongoing star formation. The lack of significant H absorption suggests that the galaxy does not belong to the E+A galaxy class (Dressler & Gunn, 1983) in which TDEs have been known to be overrepresented (Arcavi et al., 2014; French et al., 2016). We identify strong absorption lines at 4303 (G-band), 5172 (Mg i , which indicates an old stellar population), 5284 (Fe ii) and the Na i D doublet at 5890+5895 Å. Finally, the AllWISE color –=0.04 (Cutri & et al., 2014) further indicates that the black hole is most likely inactive (e.g. Wu et al. 2012; Stern et al. 2012).
To measure the galaxy mass and star formation rate (SFR), we model the spectral energy distribution (SED; see Table 1) with the software package LePhare version 2.2 (Arnouts et al., 1999; Ilbert et al., 2006)222http://www.cfht.hawaii.edu/~arnouts/LEPHARE/lephare.html. This also allows us to synthesise the host galaxy brightness in the Swift bands, which we use to subtract the host galaxy contribution from the TDE lightcurves. We generate templates based on the Bruzual & Charlot (2003) stellar population synthesis models with the Chabrier initial mass function (IMF; Chabrier, 2003). The star formation history (SFH) is approximated by a declining exponential function of the form e, where is the age of the stellar population and the e-folding time-scale of the SFH (varied in nine steps between 0.1 and 30 Gyr). These templates are attenuated with the Calzetti attenuation curve (varied in 22 steps from to 1 mag; Calzetti et al., 2000). LePhare accounts for the contribution from the diffuse gas (e.g. \ionHii regions) following the relation between SFR and the line fluxes presented in Kennicutt (1998).
From the best fit template spectrum, we derive a host galaxy stellar mass of log(M/M)=10.2, and a SFR and intrinsic consistent with 0. Using an empirical bulge-to-total (B/T) ratio (Stone et al., 2018) of 0.47 (very similar to the ratio of the PSF to Petrosian -band flux of 0.57) for this galaxy mass, we find a SMBH mass of 2 M using the M–M relation (Häring & Rix, 2004). We synthesise photometry in the Swift UVOT filters, which can be found in Table 1, to perform the host subtraction.
|GALEX NUV||21.91 0.4|
2.2 Swift X-ray and UV/optical observations
Swift’s (Gehrels et al., 2004) UltraViolet/Optical Telescope (UVOT; Roming et al. 2005) and the X-Ray Telescope (XRT; Burrows et al. 2005) started monitoring AT 2018fyk on MJD 58 383.7, approximately 8 days after the classification spectrum was taken and 14 days after the reported discovery (Brimacombe et al., 2018) by the ASAS–SN survey. Between 2018 September 22 and 2019 January 8, 52 monitoring observations were made with an average observing cadence varying between 2 and 4 days. Swift could not observe the source after 2019 January 8 due to Sun pointing constraints. We removed two observations (obsIDs: 00010883004 and 00010883038) from further analysis as they had limited XRT exposure ( 10 s) and lacked UVOT data. Figure 1 shows an X-ray image of AT 2018fyk’s field of view as observed with Swift/XRT.
The XRT observations were all performed in photon counting (PC) mode, and were reduced using the latest version of the Swift xrtpipeline provided as part of Heasoft 6.25 analysis package. Source counts were extracted using a circular aperture with a radius of 47, and corrected for the background contribution using an annulus with an inner and outer radius of 70 and 250, respectively.
We note that no source is detected in archival ROSAT observations down to a limit of 510 cts s (Boller et al., 2016). Using the webPIMMS tool333https://heasarc.gsfc.nasa.gov/cgi-bin/Tools/w3pimms/w3pimms.pl, this corresponds to a flux limit of 510 erg cm s (0.3–8 keV, assuming a power law model with n=2 typical for AGN), which translates to an upper limit to the host X-ray luminosity of 510 erg s (a blackbody model with kT=0.1 keV results in an upper limit of 1.510 erg s).
We used the uvotsource task to construct UVOT light curves, using a 5 aperture in all filters to estimate the source brightness. Background levels were estimated by using a circular region with radius of 50 centered on a nearby empty region of sky.
2.3 Optical spectroscopy
Optical spectroscopic observations were obtained with the New Technology Telescope (NTT) located at La Silla, Chile using the ESO Faint Object Spectrograph and Camera (EFOSC2; Buzzoni et al. 1984) instrument with the gr11 and gr13 grisms in combination with a 1 arcsec slit. All observations were obtained as part of the ePESSTO program. We present the observing log including observing dates, setups and exposure times in Table 2.
|Grating||Obs date||MJD||Seeing||Exposure time|
We reduce the spectroscopic data with iraf. Standard tasks such as a bias subtraction and flat field correction are performed first, after which we optimally extract the spectra (Horne, 1986) and apply a wavelength calibration using HeAr arc lamp frames. The typical spectral resolution obtained with the gr11 and gr13 setups and a 1 arcsec slit for slit-limited observing conditions is R 250 and 190 at 4000 Å, respectively (but see Table 2 for the average conditions of each observation; in seeing-limited conditions the resolution increases linearly with the average seeing). Standard star observations are used to perform the flux calibration and correct for atmospheric extinction. Given that the Galactic extinction along the line of sight is negligible, we do not try to correct for this effect. Finally, a telluric correction based on the standard star observations is applied to remove atmospheric absorption features. This is particularly useful to remove the 6800 Å absorption features located in the blue wing of the H emission line profile. Multiple spectra taken on the same night are averaged, with weights set to the overall SNR ratio between the spectra. The spectra taken on 2018–12–16 and 2018–12–17 are also averaged due to the relatively low SNR of individual exposures. The resulting spectra are shown in Figure 2, where the flux levels have been scaled to improve the readability of the plot.
2.4 Radio observations
We observed AT 2018fyk with the Australia Telescope Compact Array (ATCA) over three epochs between 2018 September 19 and 2018 November 22, under program code C3148. The observations were taken in the 750C, 6A, and 6B configurations, respectively (see Table 3). While all three are east-west configurations, the former has the inner five antennas at a maximum baseline of 750 m, with the sixth antenna located some 4.3 km away. This isolated antenna was therefore not used when imaging the first epoch, due to the possibility of artifacts arising from the large gap in uv-coverage. In all cases we observed in the 15-mm band, using two 2048-MHz frequency chunks (each comprising 2048 1-MHz channels) centred at 16.7 and 21.2 GHz. We used the standard calibrator PKS 1934-638 (Bolton et al., 1964) as a bandpass calibrator and to set the flux density scale. To solve for the time-dependent complex gains, we used the extragalactic calibrator source QSO B2311-452 (4.23 away; Veron-Cetty & Veron, 1983) in the first epoch, and QSO B2227-445 (3.29 away; Savage et al., 1977) for the two subsequent epochs, as appropriate for the relevant array configurations. We used the Common Astronomy Software Application (casa v.5.1.2; McMullin et al., 2007) package for both calibration and imaging of the data, applying standard procedures for ATCA data reduction. AT 2018fyk was not detected in any of the three epochs, with upper limits as given in Table 3.
We present the X-ray and (host subtracted) UV/optical lightcurve obtained with Swift in Figure 3. After an initial decline, the UV/optical appears to turn over around 40 days after discovery to a near constant luminosity. This plateau lasts for nearly 50 days, before the UV/optical lightcurve breaks again to start declining, while the X-rays increase in brightness.
While the flare is still more than 2.5 mag brighter than the host in the UV bands during the last Swift epoch, emission in the - and -bands was significantly contaminated by the host galaxy light even at the earliest epochs.
3.1 SED analysis
We constrain the luminosity, temperature and radius evolution of AT 2018fyk by fitting a blackbody to the Swift UVOT SED at each epoch. Due to significant contamination from the host galaxy in the reddest bands ( and ), we do not include these data points. Including these bands does not alter the general results of our analysis, but leads to bad fits and unrealistic temperatures at some epochs. We therefore fit a blackbody model to the host subtracted SED consisting of the UV bands and the -band, using a maximum likelihood approach and assuming a flat prior for the temperature between 1–5 10 K. 1 uncertainties are obtained through Markov Chain Monte Carlo simulations (Foreman-Mackey et al., 2013). Using the best-fit temperature at each epoch, we integrate under the blackbody curve from EUV to IR wavelengths to estimate the bolometric UV/optical luminosity. In addition, we also derive the characteristic emission radius at each epoch. We present the integrated UV/optical luminosity, temperature and radius evolution in Figure 4, both the epoch measurements and a 15 day binned evolution for clarity. We find a peak luminosity of 3.00.510 erg s, which declines by a factor of 4 over the first 120 days of the flare evolution. The temperature appears roughly constant initially, but there is evidence for an increase at later epochs similar to ASASSN–15oi (Holoien et al., 2016a) and AT2018zr (Holoien et al., 2018; van Velzen et al., 2019). The radius, on the other hand, stays constant for the first 70 days at 4.20.410 cm, after which it decreases by a factor of 2 in the span of 50 days. Integrating over the period with Swift coverage, we find a total UV/optical energy release of E 1.410 erg, with the uncertainties dominated by the host subtraction (the observed energy radiated at X-ray wavelengths is 10 erg).
These values are all typical when compared to the UV/optical sample of known TDEs (e.g. Hung et al. 2017; Wevers et al. 2017; Wevers et al. 2019; Holoien et al. 2018). We can also estimate the mass accretion rate by assuming a typical efficiency factor of =0.1 to convert the energy released by accretion to luminosity. Assuming that all the energy at UV/optical wavelengths is reprocessed accretion power, the accretion rate at peak can be estimated to be 0.053 M yr. For comparison, this is a factor 25 higher than the value estimated by Blagorodnova et al. (2017) for iPTF–16fnl, although the black hole mass also appears to be an order of magnitude larger for AT 2018fyk.
3.2 X-ray evolution
AT 2018fyk belongs to a growing sample of UV/optical detected TDEs observed to be X-ray bright at early times, together with ASASSN–14li (Holoien et al., 2016b), ASASSN–15oi (Holoien et al., 2016a) and PS18kh/AT2018zr (Holoien et al., 2018; van Velzen et al., 2019). In addition the source XMMSL1 J0740 (Saxton et al., 2017) was also UV/optical and X-ray bright, although it was detected in X-rays first.
The Swift XRT lightcurve shows variability of a factor 2–5 on a timescale of days, whereas the UV/optical lightcurve appears more smooth. During the first epoch the source shows a L/L ratio 220, similar to that of ASASSN–15oi (Gezari et al., 2017) and AT2018zr (Holoien et al., 2018; van Velzen et al., 2019). The X-ray emission then brightens by a factor of 10 in 6 days, and remains roughly constant for 25 days. The X-ray emission then displays a plateau similar to the UV/optical evolution, leading to a near constant L/L ratio for 70 days. Between 80 and 100 days after discovery the lightcurves decline in tandem, after which the X-rays brighten once more while the UV/optical emission keeps declining.
We fit the stacked spectrum (total exposure time of 50.2 ks, Figure 5) with a blackbody model (tbabsbbodyrad in xspec) and find a best-fit temperature (=1.12 for 178 degrees of freedom [dof]) of kT=1122 eV, negligible n and a normalisation factor norm=67585. This normalisation corresponds to a X-ray photospheric radius of R=6.90.410 cm. This in turn corresponds to the innermost stable circular orbit of an accretion disk around a non-spinning SMBH of 10 M. This is a factor of 100 lower than inferred from the bulge mass, and suggests that some obscuration (either from tidal debris or in the host galaxy) occurs. There is some hint of a soft excess below 500 eV in this model. An absorbed multi-temperature blackbody model (tbabsdiskbb) with T=1496 eV describes this soft excess slightly better (=0.98 for 178 dof). We note that an absorbed power law (phabspowerlaw) with a steep spectral slope n=4.750.16 and n=7.30.110 cm describes the data equally well (=0.94 for 178 dof). While we defer a detailed temporal X-ray spectral analysis to future work, we briefly note that there is negligible flux for emission 2 keV for most of the Swift observations. In the latest 2 epochs, however, the flux of this harder component appears to increase. Unfortunately, no more Swift observations are available after those epochs to confirm this trend. Further monitoring is required to investigate the detailed spectral evolution and the potential emergence of a harder emission component similar to XMMSL1 J0740 (Saxton et al., 2017).
3.3 Optical spectroscopy
The earliest epochs of spectroscopic observations are dominated by a hot, featureless continuum with several broad emission lines superposed. We identify broad H He ii 4686 and potentially He ii 3203 Å emission lines in the spectrum. In addition, we identify a broad emission line (or lines) in the region 3400–3600Å. This latter feature can be tentatively identified as O iii 3444 or potentially broad Fe ii (3449,3499) lines, although these identifications are uncertain.
AT 2018fyk became unobservable due to Sun constraints before the broad emission lines completely disappeared, hence we cannot perform the host galaxy subtraction in the traditional way. Instead, to identify the nature of the lines and measure their line widths and velocity offsets, we first fit cubic splines to the continuum in molly444molly is an open source spectral analysis software tool., masking all prominent emission features, host and remaining telluric absorption lines. We then subtract the continuum level to reveal the TDE emission line spectrum. Although some host contamination remains, in particular narrow absorption lines (such as the H absorption trough in the red wing of He ii), to first order this removes the featureless blackbody and host galaxy continuum contributions.
Arguably the most interesting features in the optical spectra of AT 2018fyk are the narrow emission lines that appear after the lightcurve shows a plateau in luminosity. We show the spectrum with the most prominent narrow emission features in Figure 6, including the most likely line identifications. We identify several high ionisation O iii lines, and in addition we identify several low ionisation Fe ii emission lines (ionisation potential 8 eV), particularly of the multiplets 37 and 38 with prominent features at 4512,4568,4625. We also identify low excitation He i narrow emission lines. Moreover, the increased pseudo-continuum level in Figure 7 (the green spectrum) may suggest that the emergence of these narrow Fe ii lines is accompanied by a broad component as seen in AGN, although this could also be the forest of narrow Fe ii lines that is present in the wavelength range 4300–4700 Å. This shows that the spectral diversity of TDEs is even larger than previously identified, with a class of Fe-rich events in addition to the H-, He- and N-rich TDEs (Arcavi et al., 2014; Hung et al., 2017; Blagorodnova et al., 2017; Leloudas et al., 2019).
The fact that the narrow emission lines are observed only when the lightcurve shows a plateau phase strongly suggests that they are powered by the same emission mechanism. We also note that we only see narrow emission lines in the blue part of the spectrum. Several transitions of both He i and O iii exist at longer wavelengths, and these transitions typically have stronger line strengths (for example in AGN) than the lines we observe in AT 2018fyk.
4.1 TDE classification
We classified AT 2018fyk as a tidal disruption event based on several pieces of evidence.
First, the location is consistent to within 100 pc with the nucleus of a galaxy. No signs of activity or star formation are evident from the galaxy colours and no narrow galaxy emission lines are present in the spectra, arguing against a supernova interpretation. Archival X-ray upper limits show that the X-ray emission brightened by a factor of at least 200, making an AGN flare an unlikely interpretation.
Third, the X-ray emission is an order of magnitude brighter than the brightest X-ray supernovae observed (e.g. Dwarkadas & Gruszko 2012), although rare, superluminous supernovae can produce X-ray emission similar to that observed. However, the X-ray spectra of supernovae are not expected to be well described by thermal blackbody emission.
The observed properties are broadly consistent with observed TDEs: hot (T3.510 K) UV/optical blackbody emission that does not cool over 100 days, a near-constant UV/optical colour evolution, a thermal blackbody X-ray component with a temperature of 100 eV, broad ( 15000 km s) H and He optical emission lines can all be naturally explained in the TDE scenario (Arcavi et al., 2014; Hung et al., 2017; Blagorodnova et al., 2017; Wevers et al., 2017; Holoien et al., 2018). In the remainder of this Section, we discuss several peculiar features (compared to observations of other TDEs) and how they can be explained in the TDE interpretation.
4.2 Lightcurve comparison and secondary maxima
To put the lightcurve shape into context, in Figure 8 we compare the UVW2 lightcurve of AT 2018fyk with other TDE (candidates). The -band (observed) peak absolute magnitude is V = –20.7. While the decline is monotonic for the first 40 days, similar to the TDEs ASASSN–14li and ASASSN–15oi, the lightcurve plateaus before declining at a rate similar to ASASSN–14li. This is reminiscent of the behaviour seen in the TDE candidate ASASSN–15lh, which shows a similar (albeit much more pronounced and much longer) secondary maximum in its lightcurve. Given the much higher redshift of the latter source, we also show its UVM2 lightcurve, which probes similar rest wavelengths to the UVW2 filter for the other events. The lightcurve of PS16dtm, which has been claimed to be a TDE in a NLS1 galaxy (Blanchard et al., 2017), shows a plateau phase but not the characteristic decline from peak leading up to it, as seen in AT 2018fyk and ASASSN–15lh. The UVW2 lightcurve of XMMSL1 J0740 also shows a similar, though less pronounced, rebrightening phase at 150 days (Saxton et al., 2017). For the TDE candidate ASASSN–15lh (but see e.g. Dong et al. 2016; Bersten et al. 2016 for an extreme supernova interpretation), Leloudas et al. (2016) propose that the rebrightening can be explained by taking into account the SMBH mass, which is by far the most massive of the TDE sample ( 10 M). As a consequence, all orbital pericenters become relativistic, even for shallow (low =R/R) stellar encounters. In this case, we propose a deeply penetrating (high ) event at the origin of the relativistic pericenter, which leads to two peaks in the lightcurve (Ulmer, 1999); the first maximum due to shock energy released during stream self-intersections, and the second after disk formation, powered by accretion onto the SMBH.
While we do not have an accurate black hole mass measurement for the host of AT 2018fyk, the stellar population synthesis suggests that the galaxy stellar mass is typical of other TDEs. If the SMBH mass is indeed more typical of TDE hosts, 10–10 M (Wevers et al., 2017; Wevers et al., 2019), relativistic pericenters can still occur if the encounter is deeply penetrating (high ). Below we will argue that the second peak in the lightcurve is indeed powered by efficient reprocessing of energetic photons from the central source into UV/optical emission, similar to the scenario proposed by Leloudas et al. (2016). For AT 2018fyk, however, the relativistic pericenter would be due to a high event rather than a high SMBH mass. The fact that this could be the first deeply plunging TDE is consistent with the exponential suppression of high events, expected if the main TDE seed mechanism is two-body relaxation (Stone & Metzger, 2016). For such high events, disk formation is expected to be prompt (Dai et al., 2015; Guillochon & Ramirez-Ruiz, 2015), as evident from the L/L evolution (Section 4.5). Unfortunately, the viscous timescale depends sensitively on the disk scale height and viscosity , and the impact parameter (as well as on M), which makes constraining difficult without additional constraints on the accretion disk parameters.
As a final note, identifying a TDE with more typical TDE host galaxy parameters (Wevers et al., 2019) but observational characteristics similar to ASASSN–15lh argues in favour of the TDE interpretation of that event (as opposed to a unique SN interpretation). In this interpretation the UV/optical emission and the emergence of X-ray emission after an initial non-detection are explained by the delayed formation of an accretion disk, as has been proposed for other TDEs (Gezari et al., 2017). These similarities and the natural link between the UV/optical and X-ray emission strengthen the classification of ASASSN–15lh as a TDE (Leloudas et al., 2016; Margutti et al., 2017).
4.3 Detection of low ionisation, narrow emission lines
The He ii 4686 emission line is non Gaussian in several of the spectra. Comparing its FWHM 28000 km s with that of the other lines, which range between 10–15 km s, it is hard to explain why this line is almost twice as broad if it originates in roughly the same physical region. Moreover, the line develops a distinct asymmetric blue shoulder during its evolution (Figures 7 and 9). This suggests, as has been noticed in other TDEs (e.g. Arcavi et al. 2014; Holoien et al. 2016a; Leloudas et al. 2019) that instead this line might be a superposition of several emission features. Holoien et al. (2016a) suggested that part of this line might be explained by He i 4472 in ASASSN–15oi; for AT 2018fyk the line would be redshifted by 2500 km s, which is not observed for H and He ii. While Leloudas et al. (2019) explain the asymmetry in some TDEs as a consequence of Bowen fluorescence lines, we do not observe the characteristic N iii 4097,4103 feature that is expected in this case. This suggests that in AT 2018fyk and potentially other TDEs such as ASASSN–15oi (see fig. 4 in Leloudas et al. 2019), Bowen fluorescence lines do not provide a satisfactory explanation.
The emergence of the narrow spectral lines in AT 2018fyk (Figure 6) allows us to identify the emission in this blue shoulder as Fe ii multiplet 37,38 emission lines. These are the strongest optical Fe ii multiplet lines, although depending on the excitation mechanism one might also expect emission in the NIR around 1m (Marinello et al., 2016), which is unfortunately not covered by our spectra. Given the similarity of the line profiles, we propose that the origin of the blue bump near He ii 4686 in the other two events shown in Figure 9, ASASSN–15oi and PTF–09ge, is likewise Fe ii emission, making these events part of an Fe-rich class of TDEs. We have also included the coronal line emitter and TDE candidate SDSS J0748 (Yang et al., 2013) because the line shape is remarkably similar.
These low ionisation lines have been detected in AGNs (e.g. Lawrence et al. 1988; Graham et al. 1996), with EWs that can exceed those of He ii 4686. Although the excitation mechanism(s) in AGN is somewhat ambiguous, photo ionisation (Kwan & Krolik, 1981), Ly resonance pumping (Sigut & Pradhan, 1998) and collisional excitation (depending on the particle density) have all been proposed to contribute to some extent to produce these transitions (Baldwin et al., 2004). Their strength is closely associated with the Eddington fraction in AGN (Boroson & Green, 1992; Kovačević et al., 2010). While the narrow Fe ii lines are thought to originate from a well defined region in between the broad line region (BLR) and narrow line region (NLR), the emission region of the broad component is not currently well constrained (Dong et al., 2011). One possibility is that it originates from the surface of the AGN accretion disk (Zhang et al., 2006); further evidence for an origin in the accretion disk comes from cataclysmic variables (e.g. Roelofs et al. 2006). Interestingly, Dong et al. (2010) showed that while optical Fe ii emission is prevalent in type 1 AGN, it is not observed in type 2 AGN. This suggests that the emission region is located within the obscuring torus.
More generally, the emission region is likely a partially ionised region, where the ionising photons come from a central X-ray source (Netzer & Wills, 1983). Incidentally, some of the strongest optical Fe ii lines are observed in narrow-line Seyfert 1 (NLS1) galaxies (e.g. Osterbrock & Pogge 1985), which are typically characterised by a significant soft X-ray excess below 1.5-2 keV, rapid X-ray flux and spectral variability (see e.g. the review by Gallo 2018) and potentially accreting at high fractions of their Eddington rate (Rakshit et al., 2017). Another interesting resemblance is their preferred black hole mass range, which is 10 M for both TDEs and NLS1s (Peterson, 2011; Berton et al., 2015; Chen et al., 2018). These properties are all remarkably similar to those expected/observed for TDEs.
In particular, the TDE candidate PS16dtm was suggested to be a TDE in an active galaxy (Blanchard et al., 2017); the spectrum resembles that of NLS1 galaxies, showing several optical Fe ii lines (Figure 6). PS1-10adi, another TDE candidate in an AGN, was also observed to produce transient Fe ii optical emission at late times (Kankare et al., 2017); similar features were also observed in the TDE candidates and extreme coronal line emitters SDSS J0748 and SDSS J0952 (Wang et al., 2011; Yang et al., 2013). These events all occurred around active black holes, so establishing their TDE nature is ambiguous. The resemblance of AT 2018fyk to some of these events shows that stellar disruptions can create (temporary) circumstances very similar to those in NLS1 AGN even around dormant SMBHs.
Unlike the high ionisation narrow lines such as O iii (which are thought to form within the ionisation cone of the central X-ray source in AGN), Fe ii emission requires an obscuring medium with significant particle density and optical depth as well as heating input into the gas. The presence of these Fe ii lines in the spectra of AT 2018fyk indicates that at least part of the gas is optically thick, while the X-ray spectrum shows that a bright, soft X-ray source is present, making the conditions in this TDE similar to that in NLS1 nuclei.
We inspect publicly available spectra of other TDEs, and find that the presence of narrow Fe ii lines is not unique to AT 2018fyk. We identify similar emission lines consistent with the same Fe ii multiplet 37,38 lines in optical spectra of ASASSN–15oi at late phases (330 days after discovery; Figure 6). Upon further investigation, the L/L ratio of both sources is nearly constant while the narrow lines are present (Figure 10; see also Section 4.5). This suggests that the L/L ratio evolution in ASASSN–15oi at late times may be similarly regulated by reprocessing of soft X-ray radiation in optically thick gas, analogous to the situation in AT 2018fyk and AGNs.
The formation of an accretion disk that radiates in soft X-rays, which subsequently partially ionise high density, optically thick gas surrounding the SMBH delivered by the disruption can explain the emergence of the Fe ii emission lines. At the same time, the reprocessing of X-ray, Ly and/or EUV photons can power the plateau phase in the lightcurve, explaining both peculiar features in the TDE scenario. van Velzen et al. (2018) showed that the late-time plateau phase can be explained by UV disk emission, and this can also contribute to the plateau phase seen in AT 2018fyk.
While high temporal coverage in X-ray and UV/optical wavelengths is available for only a few candidates, the UV/optical lightcurve shape of AT 2018fyk is unique among UV/optical bright TDEs. If we are indeed witnessing the assembly of an accretion disk and reprocessing of disk X-ray radiation, this implies that it does not occur with a similar efficiency in most TDEs. The first 40 days of the lightcurve, however, show typical behaviour as observed in nearly all UV/optical TDEs (Figure 8). The plateau represents an additional emission component superposed on the contribution responsible for the initial decline from peak. van Velzen et al. (2018) showed that such a secondary maximum is observed in nearly all TDEs, but several years after disruption rather than several months as observed in AT 2018fyk and ASASSN–15lh.
4.4 Broad iron emission lines?
In terms of velocities, the He ii 4686 and H lines follow a similar trend, being consistent with their respective rest wavelengths in early epochs but becoming more blueshifted up to about 2000 km s, with a blueshift of 1000 km s in the latest spectrum. Although the He ii 3202 line can tentatively be identified in the spectra, it is on the edge of the spectrum and a sudden decrease in instrumental through-put may instead be responsible for this feature. More interestingly, the (broad) line that we tentatively identify as O iii at 3444 Å or He i at 3446 Å seems to be systematically redshifted by 2000–3000 km s. Fitting a single Gaussian profile to this line, we find central wavelengths ranging between 3375 and 3500 Å during the evolution. However, the line has a rather boxy profile instead of being well described by a Gaussian. In this wavelength range, two narrow emission features with rest wavelength of 3449 and 3499 Å are visible during the nebular phase (Figure 2). While the former is consistent with either O iii 3444 Å or He i at 3447 Å, the identification of the latter is 3499 Å line is less secure. As an alternative, the NIST Atomic Spectra Database shows several strong Fe ii transitions corresponding to wavelengths close to 3449 and 3499 Å. If these line identifications as Fe ii are correct, this provides unambiguous evidence for broad Fe ii emission lines in the early spectroscopic observations (Figure 2).
We also tentatively identify the emergence of a broad emission feature around He i 5876Å that is present in several epochs. Without a solid host galaxy subtraction, however, this feature must be interpreted with caution as it is unclear what constitutes the continuum level, given the many broad features and bumps present in the spectra. In addition, there is a deep absorption feature that distorts the line shape. We tentatively identify this feature as He i 5876Å but a proper host galaxy subtraction is needed to study the line evolution in more detail.
4.5 Optical to X-ray ratio evolution
AT 2018fyk is only the third TDE with contemporaneous bright UV/optical and X-ray emission that has been observed by Swift with high cadence at both wavelength regimes. We show the ratio of integrated UV/optical luminosity to X-ray luminosity in Figure 10, where we also overplot these ratios for ASASSN–14li, 15oi and AT2018zr (Gezari et al., 2017; van Velzen et al., 2019). The L/L ratio of this source is 1 for 400 days, with some hint of a potential increase at later times. On the other hand, the evolution of the L/L ratio of ASASSN–15oi is markedly different, and has been interpreted as the delayed formation of an accretion disk (Gezari et al., 2017). The evolution of AT 2018fyk appears to broadly follow that of ASASSN–15oi, as it decreases over time. However, rather than a monotonic decrease sudden changes are apparent at early times and during the 2 most recent Swift observations. The L/L ratio appears to plateau for 80 days similar to the UV/optical lightcurves, after which it decreases as the X-ray luminosity brightens and the X-ray spectrum becomes harder.
During this plateau phase, both the X-ray and UV/optical luminosity increase in tandem (compared to the initial decline) while narrow optical emission lines corresponding to He i and both permitted and forbidden transitions of O iii appear in the spectrum. Given the high ionisation potential (higher than 35 eV), these nebular lines O iii lines typically only appear in the presence of a strongly ionising radiation field and relatively low densities. The absence of these lines in the early phases of the flare suggest that the ionising source was much fainter at those times. A scenario where we are witnessing the formation of an accretion disk during the Swift observations can explain the nebular lines if the disk radiation ionises debris (most likely the bound material, as the lines are observed at their rest wavelengths) from the disrupted star. The plateau in the lightcurve can then be explained as reprocessing of X-ray radiation into UV/optical photons, creating the right conditions for line emission. The disappearance of the nebular lines after the plateau indicate that the emitting layer of material has become fully ionised and optically thin to the X-ray radiation, which can explain the up-turn in the XRT lightcurve while the UV/optical emission becomes fainter.
4.6 Radio upper limits
We can use the radio non-detections to constrain the presence of a jet/outflow similar to that observed in ASASSN–14li (van Velzen et al., 2016a; Alexander et al., 2016; Romero-Cañizales et al., 2016; Pasham & van Velzen, 2018). To this end, we assume that the scaling relation between the radio and X-ray luminosity of a tentative jet/outflow is similar to that of ASASSN–14li, L L (Pasham & van Velzen, 2018). From Table 4 we see that the observations can marginally rule out that such a jet was produced.
If the X-ray-radio jet coupling was similar to that seen in ASASSN–14li, the difference in jet power could be explained by either a difference in available accretion power for the jet to tap into (assuming a similar jet efficiency), or by a difference in the conversion efficiency from accretion power to jet power (Pasham & van Velzen, 2018). While the latter is hard to test observationally, our observations disfavour the former scenario as the UV/optical and X-ray lightcurve and L/L evolution can potentially be explained by a relatively high (relativistic pericenter) TDE. Dai et al. (2015) have shown that this leads to higher accretion rates, hence this would result in a more powerful jet and more luminous radio emission if the jet power follows the mass accretion rate.
One scenario that could explain the radio non-detection is the presence of a tenuous circumnuclear medium (CNM; Generozov et al. 2017). Unfortunately, for AT 2018fyk, no strong constraints can be made. This illustrates the need for deeper radio observations to rule out the presence of a jet, even in the case of a low density CNM. Upper limits several orders of magnitude deeper than those presented here are required to rule out a jet power similar to ASASSN–14li in known TDEs.
|(days)||(erg s)||(erg s)|
We have presented and analysed multi-wavelength photometric and spectroscopic observations of the UV/optical and X-ray bright tidal disruption event AT 2018fyk. Gaia observations of the transient constrain the transient position to within 120 pc of the galaxy nucleus. The densely sampled Swift UVOT and XRT lightcurves show a peculiar evolution when compared to other well established TDEs but similar to ASASSN–15lh, including a secondary maximum after initial decline from peak. Optical spectra similarly showed peculiar features not previously identified, including both high and low ionisation narrow emission features. We show that similar features were present in archival spectra of at least one other TDE (ASASSN–15oi), but remained unidentified due to the complex line profiles of the broad emission lines. The main results from our analysis can be summarised as follows:
The X-ray and UV/optical lightcurves show a plateau phase of 50 days after an initial monotonic decline. When the UV/optical decline resumes, the X-rays instead turn over and increase in luminosity. Such a two component lightcurve is similar to that seen in ASASSN–15lh, albeit on shorter timescales. It can be naturally explained in the scenario of a TDE with relativistic pericenter, where the disk formation process is fast and efficient, resulting in this second maximum to occur 10s-100s of days rather than 1000s of days after disruption, as observed for most TDEs (van Velzen et al., 2018). Both a high black hole mass and deeply plunging stars can result in relativistic pericenters. We therefore suggest that while for ASASSN–15lh the peculiar lightcurve was due to its high M, in this case a high event can provide the right conditions to explain the lightcurve shape.
The X-ray spectra can be well described by a blackbody model with kT110 eV for the first 100 days. A multi-component blackbody model with kT150 eV and a steep power law with n=4.75 provide a similarly good description of the data. In the final two epochs of observations before the source became Sun constrained, the spectrum appears to develop a harder component above 2 keV which is not present at earlier times. Continued monitoring will reveal whether a hard power-law tail appears.
The optical spectra show broad H and He ii 4686 lines. We also tentatively identify broad Fe ii lines at 3449Å and 3499 Å. In particular the He ii 4686 line has a Gaussian FWHM significantly greater (2810 km s) than the other broad lines (10-1510 km s), suggesting it is a blend of multiple emission features.
We detect both high ionisation (O iii) and low ionisation (Fe ii) narrow emission lines. In particular the Fe ii complex near 4570 Å is unambiguously detected. We propose that this line complex can explain the asymmetric line profiles in this and several other Fe-rich TDEs (e.g. ASASSN–15oi, PTF–09ge).
The presence of low ionisation Fe ii emission lines requires optically thick, high density gas and (most likely) a strong source of ionising photons. Taken together with the lightcurve evolution, this suggests that the X-ray radiation is (partially) being absorbed and efficiently re-emitted in the UV/optical. When the gas is sufficiently ionised it becomes optically thin to the X-rays, leading to a decline in the UV/optical emission and the observed increase in X-ray luminosity.
The spectral features are remarkably similar to those seen in NLS1 AGN, as well as very similar to other TDE candidates in AGN such as the extreme coronal line emitters. This suggests a connection between all these events around AGN and AT2018fyk, which occurred in a quiescent SMBH. This strengthens the arguments in favour of a TDE interpretation for PS16dtm, the Kankare et al. (2017) events and the coronal line emitters.
We have illustrated that a wealth of information can be extracted from contemporaneous X-ray and UV/optical observations made possible by Swift and spectroscopic monitoring, and shown the importance of dense temporal coverage to map the detailed behaviour of both the X-ray and UV/optical emission in TDEs. Increasing the sample of TDEs with such coverage will almost certainly lead to the discovery of new behaviour in these enigmatic cosmic lighthouses, which in turn will reveal the detailed physics that occurs in these extreme environments. The detection of narrow emission lines highlights the need for medium/high resolution spectroscopic follow-up of TDEs to uncover the full diversity of their optical spectral appearance.
We thank Richard Saxton for sharing the Swift data of XMMSL1 J0740, and Suvi Gezari for sharing some of the data in Figure 10. TW is funded in part by European Research Council grant 320360 and by European Commission grant 730980. GL was supported by a research grant (19054) from VILLUM FONDEN. JCAM-J is the recipient of an Australian Research Council Future Fellowship (FT 140101082). PGJ and ZKR acknowledge support from European Research Council Consolidator Grant 647208. MG is supported by the Polish NCN MAESTRO grant 2014/14/A/ST9/00121. KM acknowledges support from STFC (ST/M005348/1) and from H2020 through an ERC Starting Grant (758638). MN acknowledges support from a Royal Astronomical Society Research Fellowship. FO acknowledges support of the H2020 Hemera program, grant agreement No 730970. Based on observations collected at the European Organisation for Astronomical Research in the Southern Hemisphere under ESO programme 199.D-0143. We acknowledge the use of public data from the Swift data archive. The Australia Telescope Compact Array is part of the Australia Telescope National Facility which is funded by the Australian Government for operation as a National Facility managed by CSIRO. This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement. We also acknowledge the Gaia Photometric Science Alerts Team (http://gsaweb.ast.cam.ac.uk/alerts).
- Alexander et al. (2016) Alexander K. D., Berger E., Guillochon J., Zauderer B. A., Williams P. K. G., 2016, ApJ, 819, L25
- Arcavi et al. (2014) Arcavi I., et al., 2014, ApJ, 793, 38
- Arnouts et al. (1999) Arnouts S., Cristiani S., Moscardini L., Matarrese S., Lucchin F., Fontana A., Giallongo E., 1999, MNRAS, 310, 540
- Auchettl et al. (2017) Auchettl K., Guillochon J., Ramirez-Ruiz E., 2017, ApJ, 838, 149
- Baldwin et al. (2004) Baldwin J. A., Ferland G. J., Korista K. T., Hamann F., LaCluyzé A., 2004, ApJ, 615, 610
- Bersten et al. (2016) Bersten M. C., Benvenuto O. G., Orellana M., Nomoto K., 2016, ApJ, 817, L8
- Berton et al. (2015) Berton M., et al., 2015, A&A, 578, A28
- Blagorodnova et al. (2017) Blagorodnova N., et al., 2017, ApJ, 844, 46
- Blanchard et al. (2017) Blanchard P. K., et al., 2017, ApJ, 843, 106
- Bloom et al. (2011) Bloom J. S., et al., 2011, Science, 333, 203
- Boller et al. (2016) Boller T., Freyberg M. J., Trümper J., Haberl F., Voges W., Nandra K., 2016, A&A, 588, A103
- Bolton et al. (1964) Bolton J. G., Gardner F. F., Mackey M. B., 1964, Australian Journal of Physics, 17, 340
- Boroson & Green (1992) Boroson T. A., Green R. F., 1992, ApJS, 80, 109
- Brimacombe et al. (2018) Brimacombe J., et al., 2018, The Astronomer’s Telegram, 12031
- Bruzual & Charlot (2003) Bruzual G., Charlot S., 2003, MNRAS, 344, 1000
- Burrows et al. (2005) Burrows D. N., et al., 2005, Space Sci. Rev., 120, 165
- Buzzoni et al. (1984) Buzzoni B., et al., 1984, The Messenger, 38, 9
- Calzetti et al. (2000) Calzetti D., Armus L., Bohlin R. C., Kinney A. L., Koornneef J., Storchi-Bergmann T., 2000, ApJ, 533, 682
- Cao et al. (2018) Cao R., Liu F. K., Zhou Z. Q., Komossa S., Ho L. C., 2018, MNRAS, 480, 2929
- Cenko et al. (2012) Cenko S. B., et al., 2012, ApJ, 753, 77
- Chabrier (2003) Chabrier G., 2003, PASP, 115, 763
- Chen et al. (2018) Chen S., et al., 2018, A&A, 615, A167
- Cutri & et al. (2014) Cutri R. M., et al. 2014, VizieR Online Data Catalog, p. II/328
- Dai et al. (2015) Dai L., McKinney J. C., Miller M. C., 2015, ApJ, 812, L39
- Dai et al. (2018) Dai L., McKinney J. C., Roth N., Ramirez-Ruiz E., Miller M. C., 2018, ApJ, 859, L20
- Dong et al. (2010) Dong X.-B., Ho L. C., Wang J.-G., Wang T.-G., Wang H., Fan X., Zhou H., 2010, ApJ, 721, L143
- Dong et al. (2011) Dong X.-B., Wang J.-G., Ho L. C., Wang T.-G., Fan X., Wang H., Zhou H., Yuan W., 2011, ApJ, 736, 86
- Dong et al. (2016) Dong S., et al., 2016, Science, 351, 257
- Dressler & Gunn (1983) Dressler A., Gunn J. E., 1983, ApJ, 270, 7
- Dwarkadas & Gruszko (2012) Dwarkadas V. V., Gruszko J., 2012, MNRAS, 419, 1515
- Fabricius et al. (2016) Fabricius C., et al., 2016, A&A, 595, A3
- Foreman-Mackey et al. (2013) Foreman-Mackey D., Hogg D. W., Lang D., Goodman J., 2013, PASP, 125, 306
- French et al. (2016) French K. D., Arcavi I., Zabludoff A., 2016, ApJ, 818, L21
- Gaia Collaboration et al. (2016) Gaia Collaboration et al., 2016, A&A, 595, A1
- Gaia Collaboration et al. (2018) Gaia Collaboration et al., 2018, A&A, 616, A1
- Gallo (2018) Gallo L., 2018, in Revisiting narrow-line Seyfert 1 galaxies and their place in the Universe. 9-13 April 2018. Padova Botanical Garden, Italy.. p. 34 (arXiv:1807.09838)
- Gehrels et al. (2004) Gehrels N., et al., 2004, ApJ, 611, 1005
- Generozov et al. (2017) Generozov A., Mimica P., Metzger B. D., Stone N. C., Giannios D., Aloy M. A., 2017, MNRAS, 464, 2481
- Gezari et al. (2008) Gezari S., et al., 2008, ApJ, 676, 944
- Gezari et al. (2017) Gezari S., Cenko S. B., Arcavi I., 2017, ApJ, 851, L47
- Godoy-Rivera et al. (2017) Godoy-Rivera D., et al., 2017, MNRAS, 466, 1428
- Graham et al. (1996) Graham M. J., Clowes R. G., Campusano L. E., 1996, MNRAS, 279, 1349
- Greiner et al. (2000) Greiner J., Schwarz R., Zharikov S., Orio M., 2000, A&A, 362, L25
- Guillochon & Ramirez-Ruiz (2015) Guillochon J., Ramirez-Ruiz E., 2015, ApJ, 809, 166
- Guillochon et al. (2014) Guillochon J., Manukian H., Ramirez-Ruiz E., 2014, ApJ, 783, 23
- Häring & Rix (2004) Häring N., Rix H.-W., 2004, ApJ, 604, L89
- Hills (1975) Hills J. G., 1975, Nature, 254, 295
- Hodgkin et al. (2013) Hodgkin S. T., Wyrzykowski L., Blagorodnova N., Koposov S., 2013, Philosophical Transactions of the Royal Society of London Series A, 371, 20120239
- Holoien et al. (2016a) Holoien T. W. S., et al., 2016a, MNRAS, 455, 2918
- Holoien et al. (2016b) Holoien T. W. S., et al., 2016b, MNRAS, 463, 3813
- Holoien et al. (2018) Holoien T. W. S., et al., 2018, arXiv e-prints, p. arXiv:1808.02890
- Horne (1986) Horne K., 1986, PASP, 98, 609
- Hung et al. (2017) Hung T., et al., 2017, ApJ, 842, 29
- Hung et al. (2019) Hung T., et al., 2019, arXiv e-prints,
- Ilbert et al. (2006) Ilbert O., et al., 2006, A&A, 457, 841
- Jansen et al. (2001) Jansen F., et al., 2001, A&A, 365, L1
- Jiang et al. (2016) Jiang N., Dou L., Wang T., Yang C., Lyu J., Zhou H., 2016, ApJ, 828, L14
- Kankare et al. (2017) Kankare E., et al., 2017, Nature Astronomy, 1, 865
- Kennicutt (1998) Kennicutt Jr. R. C., 1998, ARA&A, 36, 189
- Kochanek et al. (2017) Kochanek C. S., et al., 2017, PASP, 129, 104502
- Komossa & Bade (1999) Komossa S., Bade N., 1999, A&A, 343, 775
- Kostrzewa-Rutkowska et al. (2018) Kostrzewa-Rutkowska Z., et al., 2018, MNRAS, 481, 307
- Kovačević et al. (2010) Kovačević J., Popović L. Č., Dimitrijević M. S., 2010, ApJS, 189, 15
- Kwan & Krolik (1981) Kwan J., Krolik J. H., 1981, ApJ, 250, 478
- Law et al. (2009) Law N. M., et al., 2009, PASP, 121, 1395
- Lawrence et al. (1988) Lawrence A., Saunders W., Rowan-Robinson M., Crawford J., Ellis R. S., Frenk C. S., Efstathiou G., Kaiser N., 1988, MNRAS, 235, 261
- Leloudas et al. (2016) Leloudas G., et al., 2016, Nature Astronomy, 1, 0002
- Leloudas et al. (2019) Leloudas G., et al., 2019, arXiv e-prints,
- Lindegren et al. (2018) Lindegren L., et al., 2018, A&A, 616, A2
- Liu et al. (2017) Liu F. K., Zhou Z. Q., Cao R., Ho L. C., Komossa S., 2017, MNRAS, 472, L99
- Loeb & Ulmer (1997) Loeb A., Ulmer A., 1997, ApJ, 489, 573
- Margutti et al. (2017) Margutti R., et al., 2017, ApJ, 836, 25
- Marinello et al. (2016) Marinello M., Rodríguez-Ardila A., Garcia-Rissmann A., Sigut T. A. A., Pradhan A. K., 2016, ApJ, 820, 116
- Mattila et al. (2018) Mattila S., et al., 2018, Science, 361, 482
- McMullin et al. (2007) McMullin J. P., Waters B., Schiebel D., Young W., Golap K., 2007, in Shaw R. A., Hill F., Bell D. J., eds, Astronomical Society of the Pacific Conference Series Vol. 376, Astronomical Data Analysis Software and Systems XVI. p. 127
- Metzger & Stone (2016) Metzger B. D., Stone N. C., 2016, MNRAS, 461, 948
- Netzer & Wills (1983) Netzer H., Wills B. J., 1983, ApJ, 275, 445
- Osterbrock & Pogge (1985) Osterbrock D. E., Pogge R. W., 1985, ApJ, 297, 166
- Pasham & van Velzen (2018) Pasham D. R., van Velzen S., 2018, ApJ, 856, 1
- Peterson (2011) Peterson B. M., 2011, in Narrow-Line Seyfert 1 Galaxies and their Place in the Universe. p. 32
- Phinney (1989) Phinney E. S., 1989, in Morris M., ed., IAU Symposium Vol. 136, The Center of the Galaxy. p. 543
- Piran et al. (2015) Piran T., Svirski G., Krolik J., Cheng R. M., Shiokawa H., 2015, ApJ, 806, 164
- Planck Collaboration et al. (2014) Planck Collaboration et al., 2014, A&A, 571, A16
- Rakshit et al. (2017) Rakshit S., Stalin C. S., Chand H., Zhang X.-G., 2017, ApJS, 229, 39
- Rees (1988) Rees M. J., 1988, Nature, 333, 523
- Roelofs et al. (2006) Roelofs G. H. A., Groot P. J., Marsh T. R., Steeghs D., Nelemans G., 2006, MNRAS, 365, 1109
- Romero-Cañizales et al. (2016) Romero-Cañizales C., Prieto J. L., Chen X., Kochanek C. S., Dong S., Holoien T. W.-S., Stanek K. Z., Liu F., 2016, ApJ, 832, L10
- Roming et al. (2005) Roming P. W. A., et al., 2005, Space Sci. Rev., 120, 95
- Roth & Kasen (2018) Roth N., Kasen D., 2018, ApJ, 855, 54
- Roth et al. (2016) Roth N., Kasen D., Guillochon J., Ramirez-Ruiz E., 2016, ApJ, 827, 3
- Savage et al. (1977) Savage A., Bolton J. G., Wright A. E., 1977, MNRAS, 179, 135
- Saxton et al. (2017) Saxton R. D., Read A. M., Komossa S., Lira P., Alexander K. D., Wieringa M. H., 2017, A&A, 598, A29
- Shappee et al. (2014) Shappee B. J., et al., 2014, ApJ, 788, 48
- Shiokawa et al. (2015) Shiokawa H., Krolik J. H., Cheng R. M., Piran T., Noble S. C., 2015, ApJ, 804, 85
- Sigut & Pradhan (1998) Sigut T. A. A., Pradhan A. K., 1998, ApJ, 499, L139
- Smartt et al. (2015) Smartt S. J., et al., 2015, A&A, 579, A40
- Stern et al. (2012) Stern D., et al., 2012, ApJ, 753, 30
- Stone & Metzger (2016) Stone N. C., Metzger B. D., 2016, MNRAS, 455, 859
- Stone et al. (2018) Stone N. C., Generozov A., Vasiliev E., Metzger B. D., 2018, MNRAS, 480, 5060
- Stoughton et al. (2002) Stoughton C., et al., 2002, AJ, 123, 485
- Strubbe & Quataert (2009) Strubbe L. E., Quataert E., 2009, MNRAS, 400, 2070
- Ulmer (1999) Ulmer A., 1999, ApJ, 514, 180
- Veron-Cetty & Veron (1983) Veron-Cetty M. P., Veron P., 1983, Astronomy and Astrophysics Supplement Series, 53, 219
- Wang et al. (2011) Wang T.-G., Zhou H.-Y., Wang L.-F., Lu H.-L., Xu D., 2011, The Astrophysical Journal, 740, 85
- Wevers et al. (2017) Wevers T., van Velzen S., Jonker P. G., Stone N. C., Hung T., Onori F., Gezari S., Blagorodnova N., 2017, MNRAS, 471, 1694
- Wevers et al. (2018) Wevers T., et al., 2018, The Astronomer’s Telegram, 12040
- Wevers et al. (2019) Wevers T., Stone N. C., van Velzen S., Jonker P. G., Hung T., Auchettl K., Gezari S., Onori F., 2019, arXiv e-prints, p. arXiv:1902.04077
- Wu et al. (2012) Wu X.-B., Hao G., Jia Z., Zhang Y., Peng N., 2012, AJ, 144, 49
- Wyrzykowski et al. (2017) Wyrzykowski Ł., et al., 2017, MNRAS, 465, L114
- Yang et al. (2013) Yang C.-W., Wang T.-G., Ferland G., Yuan W., Zhou H.-Y., Jiang P., 2013, The Astrophysical Journal, 774, 46
- Zhang et al. (2006) Zhang X.-G., Dultzin-Hacyan D., Wang T.-G., 2006, MNRAS, 372, L5
- van Velzen et al. (2011) van Velzen S., et al., 2011, ApJ, 741, 73
- van Velzen et al. (2016a) van Velzen S., et al., 2016a, Science, 351, 62
- van Velzen et al. (2016b) van Velzen S., Mendez A. J., Krolik J. H., Gorjian V., 2016b, ApJ, 829, 19
- van Velzen et al. (2018) van Velzen S., Stone N. C., Metzger B. D., Gezari S., Brown T. M., Fruchter A. S., 2018, arXiv e-prints, p. arXiv:1809.00003
- van Velzen et al. (2019) van Velzen S., et al., 2019, ApJ, 872, 198
Appendix A Swift UVOT observations