A new class of flares from accreting supermassive black holes

TL;DR

This paper reports the discovery of a new class of flares from accreting supermassive black holes characterized by sustained ultraviolet-optical emission increases, spectral features including Bowen fluorescence, and implications for understanding SMBH accretion variability.

Contribution

The study introduces a new class of SMBH-related flares with unique spectral and temporal properties, expanding the understanding of accretion variability phenomena.

Findings

Identification of a new SMBH flare class with sustained emission.

Spectral features include Bowen fluorescence linked to high-velocity gas.

Flares show little evolution over 14 months, indicating prolonged accretion events.

Abstract

Accreting supermassive black holes (SMBHs) can exhibit variable emission across the electromagnetic spectrum and over a broad range of timescales. The variability of active galactic nuclei (AGNs) in the ultraviolet and optical is usually at the few tens of per cent level over timescales of hours to weeks. Recently, rare, more dramatic changes to the emission from accreting SMBHs have been observed, including tidal disruption events, 'changing look' AGNs and other extreme variability objects. The physics behind the 're-ignition', enhancement and 'shut-down' of accretion onto SMBHs is not entirely understood. Here we present a rapid increase in ultraviolet-optical emission in the centre of a nearby galaxy, marking the onset of sudden increased accretion onto a SMBH. The optical spectrum of this flare, dubbed AT 2017bgt, exhibits a mix of emission features. Some are typical of luminous, unobscured AGNs, but others are likely driven by Bowen fluorescence - robustly linked here with high-velocity gas in the vicinity of the accreting SMBH. The spectral features and increased ultraviolet flux show little evolution over a period of at least 14 months. This disfavours the tidal disruption of a star as their origin, and instead suggests a longer-term event of intensified accretion. Together with two other recently reported events with similar properties, we define a new class of SMBH-related flares. This has important implications for the classification of different types of enhanced accretion onto SMBHs.

Figures

a presents the measured line ratios, in addition to widely used classification schemes[[38](#biba.bib38), [39](#biba.bib39), [40](#biba.bib40)] – that is, the so-called Baldwin, Phillips & Terlevich (BPT) diagram[[41](#biba.bib41)]. The emission line ratios we measure, and specifically those measured from the earliest high-resolution spectrum of AT 2017bgt taken 10 days after discovery (in the rest-frame), are consistent with being driven by ionizing radiation from a “composite” emission source, which includes radiation from an accreting SMBH (that is, an AGN) in addition to young stars (that

Equations

$${\alpha }_{\mathrm{ox}}\equiv \frac{\log (F_{\nu }[2\mathrm{keV}]/F_{\nu }[2500\mathring{\mathrm{A}}])}{\log (\nu [2\mathrm{keV}]/\nu [2500\mathring{\mathrm{A}}])}\approx -\mathrm{1.22.}$$

$$t_{\mathrm{dyn}}$$

$$t_{\mathrm{th}}$$

$$t_{\mathrm{front}}$$

$$\left(\frac{M_{\mathrm{BH}}}{10^7{M}_{\odot }}\right){\left(\frac{r}{100r_{\mathrm{g}}}\right)}^{3/2}{\left(\frac{\alpha }{0.03}\right)}^{-1}{\left(\frac{h/r}{0.05}\right)}^{-1}.$$

$$t_{\nu }$$

$$\left(\frac{M_{\mathrm{BH}}}{10^7{M}_{\odot }}\right){\left(\frac{r}{100r_{\mathrm{g}}}\right)}^{3/2}{\left(\frac{\alpha }{0.03}\right)}^{-1}{\left(\frac{h/r}{0.05}\right)}^{-2}.$$

Full text

{affiliations}

Department of Physics, ETH Zurich, Wolfgang-Pauli-Strasse 27, CH-8093 Zurich, Switzerland.

School of Physics and Astronomy, Tel Aviv University, Tel Aviv 69978, Israel.

Department of Physics, University of California, Santa Barbara, CA 93106-9530, USA.

Las Cumbres Observatory, 6740 Cortona Drive, Suite 102, Goleta, CA 93117-5575, USA.

Núcleo de Astronomía de la Facultad de Ingeniería, Universidad Diego Portales, Av. Ejército Libertador 441, Santiago, Chile

Chinese Academy of Sciences South America Center for Astronomy and China-Chile Joint Center for Astronomy, Camino El Observatorio 1515, Las Condes, Santiago, Chile.

Kavli Institute for Astronomy and Astrophysics, Peking University, Beijing 100871, China.

Harvard-Smithsonian Center for Astrophysics, 60 Garden St, Cambridge, MA 02138, USA.

Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, MS 169-224, Pasadena, CA 91109, USA.

SRON Netherlands Institute for Space Research, Sorbonnelaan 2, 3584 CA Utrecht, The Netherlands.

Department of Astrophysics / Institute for Mathematics, Astrophysics and Particle Physics, Radboud University, P.O. Box 9010, 6500GL Nijmegen, The Netherlands.

Racah Institute of Physics, Hebrew University, Jerusalem 91904, Israel.

European Southern Observatory, Casilla 19001, Santiago 19, Chile

Cahill Center for Astronomy and Astrophysics, California Institute of Technology, Pasadena, CA 91125, USA.

Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge, CB3 0HA, UK.

Warsaw University Astronomical Observatory, Al. Ujazdowskie 4, 00-478 Warszawa, Poland.

Astrophysics Science Division, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA.

Joint Space-Science Institute, University of Maryland, College Park, MD 20742, USA.

Department of Astronomy and Steward Observatory, University of Arizona, 933 N. Cherry Avenue, Tucson, AZ 85721, USA.

Department of Physics, Florida State University, 77 Chieftan Way, Tallahassee, FL 32306, USA.

Carnegie Observatories, Las Campanas Observatory, Casilla 601, La Serena, Chile.

MIT Kavli Institute for Astrophysics and Space Research, 70 Vassar Street, Cambridge, MA 02139, USA.

AT 2017bgt was discovered by the All Sky Automated Survey for Supernovae (ASAS-SN[13]) as ASASSN-17cv on 2017 February 21 in the early-type galaxy 2MASX J16110570+0234002, at $z{=}0.064$ (ref. 14; see Methods §Detection and Photometric Monitoring Observations). The long-term ASAS-SN optical data show that the total emission from the galaxy brightened by a factor of $\sim$50% over a period of about two months, with half of the rise occurring within three weeks (see Supplementary Fig. 1). Follow-up Swift observations show that the UV emission increased by a factor of $\sim$75 compared to GALEX data from 2004, reaching a luminosity of $\nu L_{\nu}({\rm NUV})\simeq 8.9{\times}10^{44}\,{\rm erg\,s}^{-1}$, and that the X-ray emission increased by a factor of ${\sim}2{-}3$ compared to ROSAT data from 1990 August (see ‘Detection and photometric monitoring’ and ’Archival multi-wavelength data’ in Methods for details on all new and archival data). The archival X-ray luminosity, of $L(2{-}10\,{\rm keV})\simeq 7{\times}10^{42}\,{\rm erg\,s}^{-1}$, and the archival UV to X-ray luminosity ratio are consistent with what is commonly observed in AGN (that is, a UV-to-X-ray spectral slope of $\alpha_{\rm ox}{\approx}-1.2$; see Methods §Archival multi-wavelength data). Archival detections in the radio (from 1998; obtained by the Very Large Array) and in the mid-infrared (from 2010; obtained with the Wide-field Infrared Survey Explorer) can be accounted for by star formation in the host galaxy. Thus, AT 2017bgt experienced a dramatic increase in its UV emission, accompanied by a smaller increase in optical and X-ray emission, sometime between 2004 and 2017.

The X-ray spectral energy distribution of AT 2017bgt, as determined from our new observations over the energy range of 0.3-24 keV, using Swift/XRT, NuSTAR, and NICER, is broadly consistent with an unobscured AGN, dominated by a power-law with a photon index of $\Gamma{\simeq}1.9$ (see ’New X-ray data’ in Methods and Supplementary Fig. 2). During the current UV-bright state, AT 2017bgt exhibits UV-to-X-ray and UV-to-optical luminosity ratios that are much larger than what is typically seen in unobscured AGN. The ratio of monochromatic UV-to-X-ray luminosities, $L_{\nu}(2500\,{\rm\AA})/L_{\nu}(2\,{\rm keV})\approx 6{\times}10^{4}$, is higher than the norm by a factor of at least $\sim$50 (i.e., $\alpha_{\rm ox}{\simeq}-1.9$; see ref. 15). The corresponding UV-to-optical ratio is $\gtrsim$3.5, which is higher than the norm by a factor of at least $\sim$5.5 (ref. 16). Over the first 14 months after discovery the UV and optical flux of AT 2017bgt have shown very limited variability (Fig. 1), with a mild decline of $\lesssim$0.7 mag (factor of 2) in the UV, and $\lesssim$0.2 mag in the optical (without removal of an unknown amount of host contamination). The X-ray emission from AT 2017bgt has been roughly constant during this period (Fig. 1a).

We obtained repeated optical spectroscopy of AT 2017bgt, starting two days after discovery. The spectra display many features typical of unobscured AGN, but also some features that have never been clearly identified in such systems before (see ’Optical spectroscopy’ in Methods and Fig. 2). Among the features typical of AGN, the spectra exhibit prominent, symmetrical, single-peaked hydrogen Balmer emission lines, with full-width at half-maximum of ${\rm FWHM}{\approx}2000\,{\rm km\,s}^{-1}$. In unobscured, broad-line AGN (i.e., quasars), such emission lines are thought to originate from partially ionized gas with densities of order $10^{10}\,{\rm cm}^{-3}$. AT 2017bgt also exhibits weaker and narrower, ${\sim}500\,{\rm km\,s}^{-1}$-wide, forbidden emission lines of [O iii] $\lambda\lambda 4959,5007$ and [N ii] $\lambda\lambda 6548,6584$, which are also common in AGN. Indeed, these forbidden narrow lines are already present in the earliest optical spectra, obtained within days from the transient detection, and indicate the presence of an AGN-like ionizing source (see Supplementary Figs. 3 and 4, and ’Optical spectroscopy’ in Methods). Given that the light-travel time to the corresponding (narrow) line-emitting region is of order $\gg$100 years[17, 18], and the fact that the archival UV and X-ray data are consistent with AGN-like continuum emission, the system was most likely harbouring an actively accreting supermassive black hole (SMBH) well before the optical flux increase that triggered the transient detection.

Assuming that the optical and X-ray continuum emission from AT 2017bgt indeed originates from the vicinity of a standard accretion flow onto a SMBH, as in AGN, and using standard AGN scaling relations, we infer a mass accretion rate of $\dot{M}{\sim}0.04{-}0.11\,M_{\odot}\,{\rm yr}^{-1}$; a broad line region (BLR) size of \hbox{{R_{\rm BLR}}}{\approx}20{-}30 light-days (${\sim}2\times 10^{-2}$ pc); and, combined with an H$\beta$ line width of ${\rm FWHM}\left({\rm H}\beta\right){\simeq}2,050,{\rm km\,s}^{-1}$, a SMBH mass of $M_{\rm BH}{\approx}1.8{\times}10^{7}\,M_{\odot}$, which in turn results in an Eddington ratio of $L/L_{\rm Edd}{\sim}0.08{-}0.21$ (see ’Determination of key SMBH properties’ in Methods for details and Supplementary Fig. 5; key measured and derived properties are listed in Supplementary Table 1). The intense UV luminosity, on the other hand, could imply $L/L_{\rm Edd}{\sim}1.4$, and further suggests that the line-emitting region does not necessarily follow standard AGN scaling relations (see ’Determination of key SMBH properties’ in Methods for details). Given the width of the Balmer lines and the high accretion rates suggested by the UV-based rough estimates of $\dot{M}$ and $L/L_{\rm Edd}$, the currently observed AGN-related properties of AT 2017bgt are consistent with those of many narrow-line Seyfert 1 (NLSy1) galaxies.

Most importantly, the optical spectroscopy of AT 2017bgt presents several strong emission features that are not seen in AGN (Fig. 2), namely the strong O iii $\lambda 3133$ emission line and the double-peaked emission feature near 4680Å, all with widths consistent with that of the broad component of the H$\beta$ emission line. The redder of the two peaks near 4680 Å coincides with the He ii $\lambda 4686$ transition, which in AGN typically exhibits line intensity ratios relative to H$\beta$ of $F({\rm He}\,\textsc{ii})/F({\rm H}\beta)\leq 0.05$, but here is seen with $F({\rm He}\,\textsc{ii})/F({\rm H}\beta)\approx 0.5$ (see ref. 16). The width and intensity of these spectral features remain roughly constant during our follow-up spectroscopic observations spanning over 14 months from discovery (Fig. 1c and Supplementary Fig. 3).

The second feature, centred at 4651.6 Å, cannot be associated with a second peak of He ii emission, originating from a disk-like configuration of the He ii-emitting gas, since the other optical and near-IR (NIR) emission lines, including several He ii transitions, are single-peaked (the NIR helium lines are also exceptionally strong compared to typical AGN; see ’NIR spectroscopy’ in Methods and Supplementary Fig. 6). This feature may instead be associated with the N iii $\lambda 4640$ emission line. Some recent studies have suggested that weak emission from this transition (sometimes noted as a “Wolf-Rayet” feature) may be present in the spectra of some TDEs[19, 20, 21], although significant line blending in those cases makes the identification less secure there. While the N iii $\lambda 4640$ line is not seen in AGN spectra (Fig. 2), it can be significantly enhanced by Bowen fluorescence, which would also produce O iii $\lambda 3133$ and other O iii lines, as seen in our spectra. In Bowen fluorescence (BF), photons emitted from the Lyman$\alpha$-like transition of He ii, at 303.783 Å, excite certain states of O iii and N iii, due to the wavelength proximity between the corresponding energy transitions[22, 23]. The excited O iii and N iii states lead to a cascade of transitions, that may be observed as emission lines in the UV-optical regime. This process, and particularly strong O iii $\lambda 3133$, N iii $\lambda 4640$, and He ii $\lambda 4686$ emission, is well established to occur in some planetary nebulae and X-ray binaries[24, 25]. It is, however, generally not seen in AGN, and only a few Seyfert galaxies were reported to have narrow ($<$1,000 km s*-1*-wide) O iii $\lambda 3133$ BF lines[26, 27]. Here we present a case where BF emission lines from the BLR are associated with a (steadily) accreting SMBH, and specifically robust identification of broad O iii BF lines, and of the N iii $\lambda 4640$ line (and indeed the N iii BF cascade) in such a system, which were first predicted decades ago[28].

Comparing the $F({\rm He}\,\textsc{ii})/F({\rm H}\beta)$ and $F({\rm N}\,\textsc{iii})/F({\rm He}\,\textsc{ii})$ line intensity ratios seen in AT 2017bgt to those predictions[28] suggests that the line-emitting gas is dense, with hydrogen number density of $n_{\rm H}\gtrsim 10^{9.5}\,{\rm cm}^{-3}$, and has a high abundance of metals – and in particular of nitrogen (likely exceeding 4 times solar). These gas densities are consistent with what is expected for the BLR in AGN, and high metallicities may be expected in extremely high luminosity and/or high $L/L_{\rm Edd}$ AGN[29]. Thus, AT 2017bgt suggests that the key missing ingredient for broad BF lines in accreting SMBHs is extremely intense UV continuum emission. However, more detailed radiative transfer calculations are required to reproduce the line ratios seen in AT 2017bgt, and to link them to the enhanced UV emission.

AT 2017bgt joins two recently reported transient events in galaxy centres that also exhibit a prominent, broad and double-peaked emission feature near 4680 Å (Fig. 3). The first event was recently claimed to be a TDE in the active nucleus of the ultra-luminous infrared galaxy F01004-2237 (ref. 30). The classification of this event as a TDE mainly relied on the association of this spectral feature with He ii $\lambda 4686$, a line also seen in a class of optical and UV-bright TDEs[3, 4], though the line profile in F01004-2237 is different from that of known TDEs (see below). Another recently reported event, OGLE17aaj, shows hints of a persistent double-peaked emission feature on top of an otherwise normal optical AGN spectrum[31, 32, 33] (M.G. et al., manuscript in preparation). In both these cases, the dramatic increase in optical continuum emission was followed by a period of rather persistent continuum and line emission, over time-scales of several months – similarly to AT 2017bgt. Moreover, the width of the Balmer emission lines in all three sources classifies them as NLSy1 AGN, which are commonly thought to be powered by highly accreting SMBHs (in terms of $L/L_{\rm Edd}$). Two other recently reported nuclear transients showed slowly-evolving light curves: PS16dtm, which was claimed to be a TDE in a NLSy1 AGN[34]; and PS1-10adi, which was interpreted as a likely peculiar kind of supernova[35]. However, these two events showed no BF-related lines in their spectra. Moreover, PS1-10adi (and similar events) show a significant flux decrease over periods of a few months, not seen in AT 2017bgt. We therefore consider these two latter events to be unrelated to the new class identified here, that is: unobscured AGN-like spectra with extremely strong UV, He ii $\lambda 4686$, N iii $\lambda 4640$, and O iii $\lambda 3133$ emission, and long-term persistence of these flare features. We caution that, given the data in hand, we cannot rule out the possibility that these events may all be driven by rather similar SMBH fueling mechanisms. Determining why some events show BF features and some do not requires more detailed modelling.

We propose that AT 2017bgt, and likely similar events, are “rejuvenated” SMBHs that experienced a sudden increase in their UV-optical emission. This, in turn, enhanced the He ii Ly$\alpha$ line and initiated the BF cascades of He ii, N iii, and O iii. The extreme UV and BF emission differentiates such events from “changing look” AGN. Given the likely light-travel timescales to the region emitting the broad Balmer lines (of order a few weeks), and the fact that they are seen in the earliest optical spectra we obtained, it is most likely that the BF emission features originate in a pre-existing BLR which was suddenly exposed to the intense ionizing UV emission whose spectral energy distribution is very different from those of normal AGN. It is therefore possible that, once the intense UV continuum emission would settle back to “normal AGN” levels, the BF features would disappear. This can be tested with on-going spectroscopic monitoring. Alternatively, the BF features may be related to a newly launched outflow, perhaps driven by the sudden increase in accretion rate (see ’Relevant mechanisms for the long-lived UV flare’ in Methods). In either case, the enhanced metallicity suggested by the strong N iii $\lambda 4640$ line may or may not be related to the fast increase in UV luminosity.

The nature of the sudden UV and optical brightening event remains open. The evidence for AGN-like activity in AT 2017bgt (and the other events in this proposed class), both before and after the UV-optical event, draws attention to processes related to (thin) accretion disks that feed SMBHs. Given the very little time evolution of the continuum and line emission in AT 2017bgt-like events, as well as their spectral properties, they are unlikely to be be driven by TDEs (at least not ones similar to those reported thus far; see Fig. 4). We discuss the relevance of the TDE interpretation, and mention several other possibly relevant mechanisms, in the Methods (’Relevant mechanisms for the long-lived UV flare’). Regardless of the nature of the dramatic increase in UV flux, the properties of AT 2017bgt, and other similar transient events in galaxy nuclei, demonstrate that prominent, a broad Bowen fluorescence N iii $\lambda 4640$ emission feature may be confused with the He ii $\lambda 4686$ feature commonly seen in TDEs. Such sources thus highlight the importance of long-term, multi-wavelength monitoring campaigns, in order to identify SMBH accretion that is driven purely by TDEs, and to distinguish it from other cases of SMBHs that experience a sudden enhancement, or indeed “re-ignition”, of their accretion. We suspect that the discovery of AT 2017bgt-like events might have been hampered so far, due to their possible association with previously-known AGN (see discussion in ref. 35). Long term monitoring of AT 2017bgt-like events, and indeed the discovery of additional events of this sort, could help illuminate their nature, as well as their role in SMBH growth.