Too Cool to follow the Local Trend: Galaxies in the Early Universe

Title: Star Formation Rate and Dynamical Mass of 108 Solar Mass Black Hole Host Galaxies at redshift 6
Authors: Chris J. Willott, Jacqueline Bergeron, and Alain Omont
First Author’s Institution: Herzberg Institute of Astrophysics

Black holes reside in the centers of most galaxies. Our Milky Way, for instance, hosts a 4×106 M black hole known as Sagittarius A. Black holes interact with their host galaxies through various feedback mechanisms and affect the properties of their host galaxies.

Black holes and Star Formation

Galaxies form stars at different rates; some are slower than the others. The rate at which galaxies create stars is known as the star formation rate or SFR. Black holes are thought to play a role in their host galaxies SFR, primarily through high-energy jets they emit which then send shock waves through their surrounding medium; whether this mechanism trigger or shut down star formation is still an open-ended question.

Black holes and Velocity Dispersion in Galaxy Bulges

Stars do not move in an orderly fashion in stellar populations. Although they are still gravitationally bound as a whole, the local motions of individual stars are random in different directions. This results in a range of stellar velocities in stellar populations. Stars in the bulges of galaxies also demonstrate this velocity spread, usually denoted as σ. A famous relation known as the MBH-σ relation is a linear relation between mass of a black hole and the velocity dispersion of the stars in the galactic bulge. It implies that the growth of black holes and the growth of galaxy bulges are related and not independent of each other. For those who are interested in learning more about velocity dispersion in stellar populations, this is an Astrobites article which explores the relation between velocity dispersion and galaxy evolution.

Black holes and Quasars

Because most structural growth happened at early times, investigating the tight correlation between black hole mass and galaxy properties at early times is important in understanding the origins of these correlations. To do this, astronomers look for the most distant sources in the Universe (that are bright enough to be observed today), and they happened to be high-redshift quasars. Quasars are high-energy sources fueled by accretion of matter onto supermassive black holes (MBH ~ 108-9 M), while high-redshift quasars just means they are formed very early on in the Universe.  For example, the highest redshift quasar known to date is at z (redshift) = 7.1, corresponding to a time less than 8 Myr after the Big Bang. Studying the properties of high-redshift quasars (z >~6) and their host galaxies will enable us to understand the interaction between black holes and their host galaxies in the early Universe.

In this paper, the authors studied how the star formation rate (SFR), velocity dispersion σ, and dynamical mass Mdyn of high-redshift quasar host galaxies depend upon their black hole accretion rates and masses. A galaxy’s dynamical mass takes into account its dark matter content and so tends to be larger than its stellar mass. Since most dark matter are thought to live in bulges, the dynamical mass of a galaxy is treated as the bulge mass. The authors took observations of two z~6 quasar host galaxies from the Canada-France High-redshift Quasar Survey (CFHQS) using the Atacama Large Millimeter Array (ALMA). ALMA is used for their observations because it has the required sensitivity to probe the SFR and dynamical masses of galaxies at high redshifts. They combined these observations with their previous study of two more high-redshift quasar hosts to give a total sample of four.

The authors discovered that high-redshift quasar hosts have rather low SFR, despite having very high black hole accretion rates. This is different than what is observed in the low-redshift Universe, where star formation rate increases similarly as black hole accretion rate, as studied in this paper. Figure 1 shows the star formation rate against redshifts, where LFIR  is used as a star formation tracer. Note the clear rise in LFIR up to a peak at z~2 followed by a decline to z=6, where the rise at z~2 is attributed to the increase in SFR in massive galaxies within this redshift range.


FIG 1 – Mean far-IR luminosity for quasars at various redshifts. The blue square is one of the quasars in this paper (the other quasar is marginally detected, and so excluded from the sample). The magenta curve is a model prediction of how LFIR varies as a function of redshift. The low-redshift quasars show about 4x increase in LFIR from z = 0.3 to z = 2.4. However, the z~6 quasar has a comparatively lower LFIR, which implies lower SFR compared to their lower-redshifts counterparts.

The authors also studied the MBH-σ and MBH-Mdyn relations of their high-redshift quasar hosts. Historically, σ is measured from galaxy bulges but since bulges are less common at high-redshifts, they used [C II] line to determine σ instead. The [C II] line is also used to determine dynamical masses of the host galaxies. Figure 2 illustrates these relations. Their quasars are distributed around the local MBH-σ relation, albeit with a much larger scatter well beyond the size of their error bars. The same can be said of their MBH-Mdyn relation.


FIG 2 – The left plot shows the MBH-σ relation for various z~6 quasars. Quasars from this paper are shown as blue squares. The black line is the local standard MBH-σ relation (Kormendy & Ho 2013) with some scatter (shaded region). The right figure shows a plot of MBH versus host galaxy dynamical mass Mdyn for z~6 quasars. Again, the black line is the local correlation with some scatter (Kormendy & Ho 2013). Quasars from this paper lie within the local relation, compared to the most massive black holes (green circle and cyan diamond points) which display large offsets.

The fact that high-redshift quasars lie on the MBH-σ relation suggests that the host galaxies have undergone substantial evolution to acquire their current high dynamical masses. Nevertheless, one wonders why this mass accumulation did not lead to a high SFR, as suggested by their low LFIR. One reason could be strong jets from the central quasar inhibiting star formation. The second reason could be that LFIR is not a good tracer of star formation at such high-redshifts. Another star formation diagnostics is L[C II], and the authors noted that using L[C II] to trace SFR would bump it up by a factor of 3 for one of their quasars. As such, they recommended higher resolution [C II] observations of their quasars to more accurately constrain the various correlations between black holes and their host galaxies at high-redshifts. Until then, their current results showed that galaxies around supermassive black holes in the early Universe are just too cool to be grouped with everybody else.

About Suk Sien Tie

I am a third year PhD student at the Department of Astronomy at The Ohio State University. I am currently working on quantitative analyses of various quasar selection methods using the Dark Energy Survey (DES) and quasar variability via microlensing.

Discover more from astrobites

Subscribe to get the latest posts to your email.


  1. In Figure 1, it appears that one quasar is the only one at high redshifts. Is the fitted line showing that the SFR decreases after z~2 also supported by theory, or is that derived from observations? I’m curious if observations of other quasars at high redshifts could change this claim.

    • Hi Jennifer, it does seem odd how one can get the magenta curve from fitting the points in the plot, especially since we only have one point at the highest redshift. This is because the magenta curve is not a best fit, but rather a predicted curve from a model. This also explains why the curve is located much lower than the data points (since if we’re fitting a curve through the data points, we would expect the curve to go through the points 🙂 ). So what figure 1 intended to show is that the points seem to follow the curve relatively well, even to high redshift (that lonesome blue square point), ie observations seem to match with model.

  2. Why are bulges less common at high redshift? Are they really less frequent or just more difficult to observe/resolve?

    • Hi Sam, good question! I am not entirely sure of this, but I think it might be because the number density of massive galaxies are dropping off rapidly at high redshifts (esp. after z > 2; in fact the drop in the number density of quasars after the quasar epoch around z~2 is precisely because of the drop in the number of massive galaxies), which then causes bulges to also be less common. In one of the papers they cited, it mentioned that galaxies formed at z > 1 are also preferentially smaller, so there might be something going on with the resolution, like you mentioned.

  3. Really cool! Regarding the jets, how exactly can they inhibit star formation? I always thought of them as promoting star formation.

    • Hi Tabata, because jets from quasars are usually very energetic, they can heat the surrounding gases to very high temperatures and also sweep gases away from their host galaxies. Gases are the basic ingredients of star formation, and we need the gases to be cold so that they can collapse. So if we have gases heated up or swept away by jets, this can inhibit star formation. 🙂

      • Makes sense! Thank you very much! 🙂

  4. I actually am on the opposite side of the coin from Tabata – I am curious to learn more about why scientists think the jets from black holes might trigger star formation. It seems counter-intuitive. Does the radiation emitted in the jets heat up or excite the clouds where star formation takes place?

    • Hi Zoey, you’re definitely warranted in your intuition that jets inhibit star formation (see reply to Tabata’s question above). Nevertheless, there are more cases of jet-induced star formation that we know, eg Centaurus A and Minkowski’s object (a peculiar star-forming galaxy). In most of these cases, the jet is weaker (ie, not moving as fast) and emits mostly in the radio. The radio emission and the weaker shock waves from these jets are thought to trigger star formation in their host galaxies. 🙂

  5. I agree with Sam here, it is possible that we are obtaining a non-random sample of high-redshift galaxies because they are more difficult to observe. This potential response bias in our sample has to be accounted for before we can make any claims as to a difference between high redshift galaxies and the local trend.

    • Hi George, definitely that when we’re doing observations of any high-redshift objects, there is a strong bias toward observing the brightest ones (in fact, this is known as the Malmquist bias). As you said, to make any meaningful claims, this bias needs to be correctly accounted for. Based on my inside knowledge (that the first author has worked on high-redshift quasars for a long time now), I believe they accounted for that in this paper.

  6. Thanks for this great post! Are there any telescopes other than ALMA that have sufficient sensitivity to to probe the SFR and dynamical masses of galaxies at high redshifts?



  1. Demasiado Frío para Seguir la Tendencia Local: Galaxias en el Universo Temprano | Astrobites en español - […] Título: Star Formation Rate and Dynamical Mass of 108 Solar Mass Black Hole Host Galaxies at redshift 6 Autores:…

Leave a Reply