This week’s paper: Cold flows and the first quasars

Meeting 7 of the Astronomy Twitter Journal Club (Thursday July 28th 2011, 19:00 UT) will discuss Cold flows and the first quasars (Di Matteo et al. July 2011). arXiv:1107.1253v1 [astro-ph.CO]

A recent discovery of a bright quasar at a red shift of about 7 means that theorists need to figure out how a billion solar mass black hole can grow within just 750 million years after the big bang. It is believed that accretion onto such black holes powers quasars.

The authors of this paper claim to have built a simulation that can create such black holes within a few hundred million years by using “cold flow accretion”, in which filaments of cold, dense gas stream onto dark matter halos. And importantly, the quasars don’t begin to heat the surrounding gas and stop the accretion until z ~ 7.

Their simulation also suggests that the competing theory for supermassive black hole formation, galaxy mergers, is not significant during the early Universe.

Those nice people at astrobites have conveniently already reviewed the paper. So, head over there to read the excellent astrobites summary.

The paper concludes:

This cold-flow driven scenario for the formation of quasars implies that they should be ubiquitous in galaxies in the early universe and that major (proto) galaxy mergers are not a requirement for efficient fuel supply and growth, particularly for the earliest SMBHs.

The paper was suggested by Marcel Haas. He writes:

This paper describes a new cosmological smoothed-particle hydrodynamics (SPH) simulation, MassiveBlack, the biggest ever done. They simulate an impressive 2×3200³ particles (dark matter and baryonic particles, respectively) in a box slightly larger than the Millennium Simulation (533 Gpc/h on a side). For the simulation to run, significant modifications had to be made to Gadget in order to run on 10⁵ cores on a new computer at NICS. The large box and number of particles make the simulation have a resolution that is fairly modest, compared to some other simulations.

In the paper they focus on the growth of super-massive black holes in the early universe. They propose a scenario in which these black holes are fed by cold streams of gas, much like star formation is supposed to be regulated by such streams in recent work (e.g. Dekel et al 2009). At halo masses higher than 10¹² M_Sun these cold streams are destroyed and quasar growth will be slower. Black holes in these first galaxies are supposed to be more massive than the local M_BH – sigma relation suggests.

Some issues regarding this paper came to mind when reading it:

• What other relevant physics is going on and how is it simulated? In particular, I’m thinking of gas cooling, star formation, supernova feedback and reionization.
• Is the resolution of these simulations sufficient to study the effects the authors are interested in? I did not see a mention of tests for numerical convergence.
• How unique is their Black Hole growth and feedback model and how does that model relate to the two questions above?

Note that there is also a related paper (same simulation, similar topic): Early Black Holes in Cosmological Simulations: Luminosity Functions and Clustering Behaviour.