News & Views | Published:


A black hole changes its feeding habits

Nature volume 540, pages 4849 (01 December 2016) | Download Citation

In the 1980s, the gas surrounding a black hole in a nearby galaxy began to emit much more radiation than before. This change has unexpectedly reversed in the past five years, questioning our understanding of these extreme phenomena.

Supermassive black holes are millions to billions of times more massive than the Sun and reside at the centres of almost all large galaxies such as the Milky Way1. Some of these black holes are actively growing by accretion, a process in which material close to the black hole cannot escape the strong gravitational field and is ultimately pulled into the black hole. These growing black holes, or active galactic nuclei (AGNs), release large amounts of energy from the accretion process, making them visible across the Universe. Writing in Astronomy & Astrophysics, McElroy et al.2 and Husemann et al.3 report that the emission spectrum of a nearby AGN has changed dramatically in the past five years, providing valuable insight into the physical processes involved in black-hole accretion.

Clues to the growth of black holes are imprinted in the optical spectra of AGNs. Just as a prism splits white light into its component colours, these spectra reveal the composition of matter, because atoms, ions and molecules emit light at specific wavelengths dictated by quantum physics. In the case of AGNs, light emitted by the accreting material energizes the surrounding gas, which produces prominent emission lines in the gas's optical spectrum.

These emission lines also tell us about the motion of gas in the AGN. Because gas close to a black hole orbits at high velocities (up to thousands of kilometres per second), the emission lines from these regions are broadened as a result of the Doppler effect: light emitted from gas moving away from Earth is shifted to longer wavelengths, whereas gas moving towards us emits light that is shifted to shorter wavelengths. Conversely, gas farther away from the black hole travels more slowly (hundreds of kilometres per second) and produces narrow emission lines. For some AGNs, we observe both broad and narrow lines, whereas in others we see narrow lines only.

The theory of AGN unification4,5 posits that all AGNs are inherently the same — namely, these phenomena are accreting supermassive black holes that are surrounded by gas and dust in the shape of a torus. Therefore, if we are looking through this enshrouding material, the gas near the black hole will be blocked from our view and we will see narrow emission lines only. However, if we are looking through the opening of the torus, we will observe both the broad and the narrow lines.

Although this theory explains why the optical spectra of many AGNs do not show broad emission lines, 'changing-look' AGNs have been discovered in which the broad lines have either disappeared or appeared over time. Some of these phenomena can be explained through the lens of AGN unification — if the obscuring material is patchy, it can block or reveal the gas close to the black hole as it orbits6. However, variable obscuration cannot explain other changing-look AGNs7,8 such as Mrk 1018, the one studied by McElroy et al. and Husemann and colleagues.

In the 1980s, the optical spectrum of Mrk 1018 transitioned from having only narrow emission lines to also showing prominent broad lines9. Variable gas clouds were ruled out as being responsible for this transition. Instead, the favoured interpretation was that the gas was energized by an enhanced rate of accretion onto the black hole (Fig. 1). Such a change could have been caused by an increase in the gas supply near the black hole. For at least one changing-look AGN, it has been postulated that a nearby star was ripped apart by the gravitational (tidal) forces of the black hole, and that this provided the fuel to reignite the AGN10.

Figure 1: The variability of Mrk 1018.
Figure 1

Black holes can grow by pulling (accreting) interstellar material from their surroundings. This accretion process energizes the gas that surrounds the black hole, producing prominent lines in the gas's emission spectrum. Because gas close to the black hole is extremely energetic, the emission lines from this region are broadened as a result of the Doppler effect. In the 1980s, prominent broad lines appeared9 in the optical spectrum of a black-hole system called Mrk 1018. The leading explanation for this change (indicated by the blue arrow) is that the accretion rate onto the black hole increased, energizing more of the gas and expanding the size of the broad-line region. McElroy et al.2 and Husemann et al.3 have combined optical and X-ray observations to show that, in the past five years, Mrk 1018 has returned to its original state (indicated by the white arrow). The authors suggest that this transition is due to a decrease in the black-hole accretion rate. Image: Adapted From Michael S. Helfenbein/Yale Univ.

In 2015, Mrk 1018 was observed using the Multi Unit Spectroscopic Explorer (MUSE) installed on the Very Large Telescope in northern Chile, as part of a programme to study nearby AGNs. Using these observations, McElroy and colleagues serendipitously discovered that the optical spectrum of Mrk 1018 has changed again, returning to the state before the broad lines became prominent. By comparing the MUSE spectra with archival data from the 1980s to 2009, McElroy et al. show that this latest transition happened in the past five years. This result constrains the lifetime of the previous stage (in which the spectra showed broad lines) to between 25 and 30 years. Like the authors of the earlier analysis9, McElroy and collaborators suggest that the transition was probably caused by a change (in this case, a decrease) in the black-hole accretion rate, rather than by variable obscuration.

In a companion paper, Husemann and colleagues focus on the X-ray properties of Mrk 1018. In AGNs, X-rays are produced in a hot 'corona' that is closer to the black hole than the gas that emits broad emission lines, providing a direct probe of black-hole accretion. Husemann et al. compare data that were taken in 2016 by two space-based X-ray observatories — Chandra and the Nuclear Spectroscopic Telescope Array (NuSTAR) — with Chandra observations from 2010. The authors find that the amount of X-rays emitted by Mrk 1018 has significantly diminished over this time, and claim that this reduction is not due to absorption of the X-rays by the gas that surrounds the accreting material. Their results therefore support the conclusion that the transition of Mrk 1018 is not caused by variable obscuration.

From the hundreds of thousands of AGNs whose existence has been confirmed using optical spectroscopy, about 20 changing-look AGNs have been discovered, and only 3, including Mrk 1018, have had broad emission lines appear and then disappear7,11. Many of the known changing-look AGNs were discovered serendipitously, which has motivated dedicated searches to find more examples, taking advantage of spectroscopic data over multiple epochs of the Sloan Digital Sky Survey12,13,14. The pace of discovery will increase in the coming decade, thanks to current and upcoming observatories such as Pan-STARRS and the Large Synoptic Survey Telescope, which will monitor the sky multiple times with the aim of detecting transient phenomena.

Although changing-look AGNs are rare, they are incredibly useful because they provide us with an opportunity to learn about the physics of accretion and the evolution of AGNs. The analyses of McElroy et al. and Husemann et al. teach us about the intermittency of black-hole accretion, suggesting that the picture is more complicated than the shutting down of an AGN. The timescale of these transitions also poses serious challenges to standard accretion theories15, which operate over periods that are seemingly too short8,10. Changing-look AGNs therefore serve as valuable tools for refining theoretical models, using observations that were not previously possible.



  1. 1.

    & Annu. Rev. Astron. Astrophys. 51, 511–563 (2013).

  2. 2.

    et al. Astron. Astrophys. 593, L8 (2016).

  3. 3.

    et al. Astron. Astrophys. 593, L9 (2016).

  4. 4.

    Annu. Rev. Astron. Astrophys. 31, 473–521 (1993).

  5. 5.

    & Publ. Astron. Soc. Pacif. 107, 803–845 (1995).

  6. 6.

    , & Astrophys. J. 410, L11 (1993).

  7. 7.

    et al. Astrophys. J. 796, 134 (2014).

  8. 8.

    et al. Astrophys. J. 800, 144 (2015).

  9. 9.

    et al. Astrophys. J. 311, 135 (1986).

  10. 10.

    et al. Mon. Not. R. Astron. Soc. 452, 69–87 (2015).

  11. 11.

    , , , & Astrophys. J. 519, L123 (1999).

  12. 12.

    et al. Mon. Not. R. Astron. Soc. 455, 1691–1701 (2016).

  13. 13.

    et al. Astrophys. J. 826, 188 (2016).

  14. 14.

    et al. Mon. Not. R. Astron. Soc. 457, 389–404 (2016).

  15. 15.

    & Mon. Not. R. Astron. Soc. 175, 613–632 (1976).

Download references

Author information


  1. Stephanie LaMassa is at NASA Goddard Space Flight Center, Greenbelt, Maryland 20771, USA.

    • Stephanie LaMassa


  1. Search for Stephanie LaMassa in:

Corresponding author

Correspondence to Stephanie LaMassa.

About this article

Publication history



Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing