I just stumbled across the following answer "update" from 2013:

This just in, recent observations have ruled out the possibility that SMBHs gain mass only through merging with other black holes. For more, see this Astronomy.com article.

Question: How do we know that supermassive black holes can gain mass by means other than merging with other supermassive black holes? I can't figure it out from the linked Astronomy.com article. In fact I can barely read it behind all the blinking advertisements that it now displays (and probably didn't back in 2013)

Potentially related:

  • 3
    $\begingroup$ Because quasars exist. I'm not sure the phrasing of your question is right. $\endgroup$
    – ProfRob
    May 16, 2021 at 7:27
  • 3
    $\begingroup$ To extend what @ProfRob said: quasars are one of the ways we discovered that SMBHs exist, and are SMBHs accreting mass by a means other than merging with other SMBHs. The question needs at least to say something like 'how do we know that most mass gain by SMBHs is not by mergers with other SMBHs?'. $\endgroup$
    – user38308
    May 16, 2021 at 11:01
  • 5
    $\begingroup$ " In fact I can barely read it behind all the blinking advertisements that it now displays " Browsing the web without adblocker? You are a lot braver than I am. $\endgroup$
    – Polygnome
    May 16, 2021 at 14:50
  • 3
    $\begingroup$ How would an SMBH avoid gaining mass when it eats a non-holey object? $\endgroup$
    – Vikki
    May 16, 2021 at 21:13
  • 1
    $\begingroup$ @uhoh For Chrome I'm using uBlock Origin, and so far I've never had any issues. I can't say anything about Safari, though. $\endgroup$
    – Polygnome
    May 17, 2021 at 17:21

3 Answers 3


The idea behind the paper (Shannon et al. 2013) that article is based on is to measure the gravitational wave background (GWB) produced by mergers of supermassive black holes, and determine which models of SMBH merger histories can replicate the SMBH population and the corresponding gravitational wave background. In this paper, measuring the GWB is done using pulsar timing - in particular, using the Parkes Pulsar Timing Array and some supplementary data from Arecibo. A pulsar timing array (PTA) detects (or attempts to detect - we have nothing definite yet) gravitational waves by measuring when individual pulses from pulsars arrive at radio telescopes. Gravitational waves should change the times of arrival (TOAs), so if you have enough TOAs from a selection of millisecond pulsars, you can constrain the GWB spectrum's amplitude $A$ and its associated energy density $\Omega_{\text{GW}}$.

The authors applied the aforementioned methodology and then considered four different models of SMBH population synthesis to try and replicate the observational results. The one which involved mass growth entirely through black hole-black hole mergers at low redshifts was ruled out at the 91% confidence level by the pulsar timing array measurements, while the others were more favored. These other models generally included SMBHs growing by accreting gas. No single one is a clear favorite, but the authors were able to rule out substantial regions of parameter space, and these three models were certainly favored more than the pure black hole-black hole merger model.

I want to emphasize something here that wasn't emphasized in the article: That merger model is focused on SMBH growth at low redshifts, and in particular on the SMBH population evolution at $z<1$. Why? It's because this is where the dominant component of the GWB we detect today is expected to come from (see e.g. McWilliams et al. 2013), so it's much easier to say something about low-redshift populations based on PTA data. This is also why the group is interested in how the other models fare at low redshifts for doing things like replicating the quasar luminosity function at $z<1$.

I feel like saying that the pure merger model has been entirely "ruled out" is maybe a bit strong; fortunately, pulsar timing arrays have made major advances since 2013, and we have stricter limits on $A$ and a better understanding of SMBH populations. The 2013 paper used only 20 pulsars, but PTAs have easily more than doubled that number in the interim, and that's only going to increase in the years to come. At the same time, the other models are certainly strongly preferred. I can look into whether this particular analysis has been replicated with newer data.

  • $\begingroup$ As a total layman I wonder: What would prevent mass from accretion disks slowly spiral into black holes, thus contributing to their mass? (I understand that mechanism would be impossible or negligible if mergers were the only mechanism.) Especially since we have evidence of very dynamic, long-lasting activity like the huge jets emanating from supermassive black holes. $\endgroup$ May 17, 2021 at 11:27

We know that black holes can gain mass other than merging with other black holes because we see high redshift quasars.

The luminosity of quasars is caused by the accretion of mass into their central black holes.

There is no question that supermassive black holes gain mass in ways other than by mergers with other black holes, because otherwise we wouldn't be able to observe them.

The main question is how some of them can get so big so fast - i.e. why we can see luminous quasars beyond $z=6$.

This is still not resolved. Mergers, the formation of intermediate mass "seed" black holes and accretion modes that circumvent the naive Eddington limit are all still possibilities afaik.


tl;dr: A study reported a possible scale (in)variance of a tidal disruption event or TDE.

Axios' Big and little black holes feed the same way has their own recent take on this question.

What they found: The study (Rapid accretion state transitions following the tidal disruption event AT2018fyk) in the Astrophysical Journal suggests all black holes go through a similar cycle when feeding whether they are about 10 times the mass of our Sun or a supermassive black hole at the center of a galaxy.

  • Scientists know that when relatively small black holes get a large helping of gas or dust, they move into a phase where the object feeds from a disc surrounding the black hole — called an accretion disk.
  • As the disc collapses, the area around the black hole can glow brightly in X-ray, and that eventually gives way to the object becoming quiet again. This all happens over the course of a few weeks to months.
  • Researchers thought this process would take too long for them to watch the whole thing play out with a supermassive black hole, but the new study found feeding can speed up if the black hole gets a large meal all at once, like when it shreds a star.

How they did it: The researchers behind the new study watched as a supermassive black hole 860 million light-years away gobbled up a star in 2018, giving them a firsthand view of how these huge black holes eat.

The bottom line: “When you throw a ball of gas at them, they all seem to do more or less the same thing," study author Dheeraj Pasham, of MIT, said in a statement. "They’re the same beast in terms of their accretion."

"We’ve demonstrated that if you’ve seen one black hole, you’ve seen them all, in a sense."

The abstract of the linked paper:

Following a tidal disruption event (TDE), the accretion rate can evolve from quiescent to near-Eddington levels and back over months - years timescales. This provides a unique opportunity to study the formation and evolution of the accretion flow around supermassive black holes (SMBHs). We present two years of multi-wavelength monitoring observations of the TDE AT2018fyk at X-ray, UV, optical and radio wavelengths. We identify three distinct accretion states and two state transitions between them. These appear remarkably similar to the behaviour of stellar-mass black holes in outburst. The X-ray spectral properties show a transition from a soft (thermal-dominated) to a hard (power-law dominated) spectral state around $L_{bol}$ ∼few ×10−2 $L_{Edd}$, and the strengthening of the corona over time ∼100--200 days after the UV/optical peak. Contemporaneously, the spectral energy distribution (in particular, the UV-to-X-ray spectral slope αox) shows a pronounced softening as the outburst progresses. The X-ray timing properties also show a marked change, initially dominated by variability at long (>day) timescales while a high frequency (∼10−3 Hz) component emerges after the transition into the hard state. At late times (∼500 days after peak), a second accretion state transition occurs, from the hard into the quiescent state, as identified by the sudden collapse of the bolometric (X-ray+UV) emission to levels below 10−−3.4 $L_{Edd}$. Our findings illustrate that TDEs can be used to study the scale (in)variance of accretion processes in individual SMBHs. Consequently, they provide a new avenue to study accretion states over seven orders of magnitude in black hole mass, removing limitations inherent to commonly used ensemble studies.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .