What is the significance of high redshift black holes for the early universe?

[wolram]

http://arxiv.org/PS_cache/astro-ph/pdf/0506/0506040.pdf

Title: Rapid growth of high redshift black holes
Authors: Marta Volonteri, Martin J. Rees
Comments: Submitted to ApJ letters. AASTeX format. 11 pages, 1 colour figure

We discuss a model for the early assembly of supermassive black holes (SMBHs) at the center of galaxies that trace their hierarchical build-up far up in the dark halo `merger tree'. Motivated by the observations of luminous quasars around redshift z=6 with SMBH masses of billion solar masses, we assess the possibility of an early phase of stable super-critical quasi-spherical accretion in the BHs hosted by metal free halos with virial temperature larger than 10000 K. We assume that the first `seed' black holes formed with intermediate masses following the collapse of the first generation of stars, in mini-halos collapsing at z=20-30 from high peaks of density fluctuations. In high redshift halos with virial temperature larger than 10000 K, conditions exist for the formation of a fat disc of gas at T_gas=5000-10000 K. Cooling via hydrogen atomic lines is in fact effective in these comparatively massive halos. The cooling and collapse of an initially spherical configuration of gas leads to a rotationally supported disc at the center of the halo if baryons preserve their specific angular momentum during collapse. The conditions for the formation of the gas disc and accretion onto a central black holes out of this supply of gas are investigated, as well as the feedback of the emission onto the host and onto the intergalactic medium. We find that even a short phase of supercritical accretion eases the requirements set by the z=6 quasars.

 

[SpaceTiger]

This is an interesting paper and, far as I can tell, completely reasonable. I don't know whether or not it's right, but I wouldn't be overly hesitant to jump explanation such as "super-accretion". Why?

Whenever one is considering things such as the "highest redshift quasars" and the limits they present, one should always keep in mind the extreme selection effects involved in the observations. In other words, the quasars wouldn't be observed if they weren't so bright, so if even a small fraction of "baby" quasars grew to large masses by z~6, they might be seen by our telescopes. Because we're selecting outliers to begin with, it's not unreasonable that their formation history might have been special as well.

Honestly, I think this theoretical bandwagon, much like the low quadrupole in the CMB, is fairly premature. We should wait until we are more completely and precisely sampling the high-z population of quasars before we conclude any kind of theoretical crisis.

 

[turbo]

The paper presumes that mass can accrete at super-Eddington rates as long as the metallicity is low.

In this paper, we have envisaged an early stage of supercritical accretion during the global evolution of a SMBH (see also Kauffmann & Haehnelt 2000). Fuel is supplied by a dense gaseous disc forming in halos with Tvir > 104K, where hydrogen atomic cooling is effective. If the disc rotates as a rigid body, at the Bondi radius of the central MBH the inflow of the gas is quasi-radial and the MBH can accept most of the infalling mass. Even if this phase lasts only until the universe in enriched by metals, the quick start allows the holes in the most massive halos to reach the high SMBH masses suggested by the SDSS quasars.

However, the distant z~6-6.5 quasars in the SDSS exhibit solar and super-solar metallicities, and no evolution of metallicity with redshift has been observed.

http://cosmos.as.arizona.edu/~thompson/pubdb/docs/barth03a.pdf
http://citebase.eprints.org/cgi-bin/fulltext?format=application/pdf&identifier=oai%3AarXiv.org%3Aastro-ph%2F0112075
http://cosmos.as.arizona.edu/~thompson/pubdb/docs/freudling03a.pdf

It is reasonable to ask why we see no evolution in metallicity z=1~6.5.

We therefore argue that this super-critical accretion phase would end when the Universe is enriched by metals at 6 < z < 10.

This sounds reasonable, but again, at what redshift will quasars exhibit sub-solar metallicities? At some redshift, we should see quasars that are feeding on a mix of highly metallized stars and gases AND poorly metallized stars and gases, and one would expect a continuum in the metallicity curve, not an abrupt discontinuity somewhere in 6<z<10. The LBT and Webb will allow us to probe these redshifts soon.

 

[Chronos]

And that is just plain wrong. There is plenty of evidence of metallicity evolution. But you won't see it when you are blinded by selection effects in your sources. How many dozens of references will it take to dispell that mythology? BTW, Wolram, nice find and very credible. Rees is top drawer in my book. I read that paper earlier today. You saved me the trouble by starting this thread.

 

[wolram]

Chronos

Maybe we should start a metallicity thread?

http://arxiv.org/PS_cache/astro-ph/pdf/0105/0105397.pdf

to start with?

 

[turbo]

And that is just plain wrong. There is plenty of evidence of metallicity evolution. But you won't see it when you are blinded by selection effects in your sources. How many dozens of references will it take to dispell that mythology? BTW, Wolram, nice find and very credible. Rees is top drawer in my book. I read that paper earlier today. You saved me the trouble by starting this thread.

I have cited these papers before, Chronos, but I will trot them out again with short quotes from their abstracts. The sample sizes are small in these studies, but you simply cannot dismiss every discordant observation as a selection effect.

http://citebase.eprints.org/cgi-bin/citations?id=oai:arXiv.org:astro-ph/0112075

We combine measurements of the CIV line and limits on the HeII emission with the NV line measurements from the optical spectra to derive line ratios, and by implication the abundances of these early quasar environments. The results are indistinguishable from those of lower redshift quasars and indicate little or no evolution in the abundances from z ~ 6 to z ~ 2. The line ratios suggest supersolar metallicities, implying that the first stars around the quasars must have formed at least a few hundreds of Myrs prior to the observation, i.e. at redshift higher than 8.

http://citebase.eprints.org/cgi-bin/citations?id=oai%3AarXiv%2Eorg%3Aastro%2Dph%2F0303424

The strength of this complex and the ratio of its integrated flux to that of Mg II lambda 2800 are comparable to values measured for QSOs at lower redshifts, and are consistent with Fe/Mg abundance ratios near or above the solar value. There thus appears to be no evolution of QSO metallicity to z~6.

http://citebase.eprints.org/cgi-bin/fulltext?format=application/pdf&identifier=oai%3AarXiv.org%3Aastro-ph%2F0311043

We find evidence of emission in the Fe II complex centered
near 2500°A (rest) in all five objects. We estimate Fe II / Mg II 2800
flux ratios comparable to those measured in QSOs at lower redshifts,
which indicate metallicities near or above the solar level. We discuss the
possible implications of this result assuming the iron enrichment to have
been produced mainly by Type Ia supernovae.

 

[Chronos]

Agreed. The sample sizes have been small, and error bars large. But that is changing and trends in metallicity evolution have been emerging:

The Age-Metallicity Relation of the Universe in Neutral Gas: The First 100 Damped Lya Systems
http://arxiv.org/abs/astro-ph/0305314

Chemical Abundances in SFG and DLA
http://arxiv.org/abs/astro-ph/0504389

Damped Lyman Alpha Surveys and Statistics - A Review
http://arxiv.org/abs/astro-ph/0505479