An expanded M_bh-sigma diagram, and a new calibration of active galactic nuclei masses

TL;DR

This paper refines the black hole mass-velocity dispersion relation using an expanded galaxy sample, accounting for biases and galaxy morphology, and recalibrates active galactic nuclei mass estimates with a new scaling factor.

Contribution

It provides an updated, bias-corrected M_bh-sigma relation for various galaxy types and introduces a revised virial scaling factor for AGN black hole mass estimates.

Findings

Steeper M_bh-sigma relation after bias correction.

Updated scaling factor f = 2.8 for AGN mass estimates.

Relations have a scatter as low as 0.34 dex.

Abstract

[Abridged] We present an updated and improved M_bh-sigma diagram containing 64 galaxies for which M_bh measurements (not just upper limits) are available. Due to new and increased black hole masses at the high-mass end, and a better representation of barred galaxies at the low-mass end, the "classical" (all morphological type) M_bh-sigma relation for predicting black hole masses is log(M_bh/M_Sun) = (5.13+/-0.34)log[sigma/200] + (8.13+/-0.05), with an rms scatter of 0.43 dex. Modifying the regression analysis to correct for a hitherto over-looked sample bias in which black holes with masses <10^6 M_Sun are not (yet) detectable, the relation steepens further to give log(M_bh/M_Sun) = (5.95+/-0.44)log[sigma/200] + (8.15+/-0.06). We have also updated the "barless" and "elliptical-only" M_bh-sigma relations introduced by Graham and Hu in 2008 due to the offset nature of barred/disc galaxies. These relations have a total scatter as low as 0.34 dex and currently define the upper envelope of points in the M_bh-sigma diagram. These relations also have a slope consistent with the value 5, in agreement with the prediction by Silk & Rees based on feedback from massive black holes in bulges built by monolithic-collapse. Using updated virial products and velocity dispersions from 28 active galactic nuclei, we determine that the optimal scaling factor f - which brings their virial products in line with the 64 directly measured black hole masses - is 2.8^{+0.7}_{-0.5}. This is roughly half the value reported by Onken et al. and Woo et al., and consequently halves the mass estimates of most high-redshift quasars. We have explored the results after separating the samples into barred and non-barred galaxies, and we have also developed a preliminary corrective term to the velocity dispersion based on bar dynamics.