Why do galaxies stop forming stars? I. The passive fraction - black hole mass relation for central galaxies

Abstract

We derive the dependence of the fraction of passive central galaxies on the mass of their supermassive black holes for a sample of over 400,000 SDSS galaxies at z < 0.2. Our large sample contains galaxies in a wide range of environments, with stellar masses 8 < log(M * /M ⊙ ) < 12, spanning the entire morphological spectrum from pure disks to spheroids. We derive estimates for the black hole masses from measured central velocity dispersions and bulge masses, using a variety of published empirical relationships. We find a very strong dependence of the passive fraction on black hole mass, which is largely unaffected by the details of the black hole mass estimate. Moreover, the passive fraction relationship with black hole mass remains strong and tight even at fixed values of galaxy stellar mass (M * ), dark matter halo mass (M halo ), and bulge-to-total stellar mass ratio (B/T ). Whereas, the passive fraction dependence on M * , M halo and B/T is weak at fixed M BH . These observations show that, for central galaxies, M BH is the strongest correlator with the passive fraction, consistent with quenching from AGN feedback.

An artificial neural network approach for ranking quenching parameters in central galaxies

“…However, bulge mass is not the best single variable found here in the ANN minimisation procedure: centralised velocity dispersion yields significantly higher AUC values (and hence tighter correlations to the star forming state of galaxies) than bulge mass. This was also argued for previously through an analysis of the passive fraction -(estimated) black hole mass relation in Bluck et al (2015), and is consistent with the importance of central density or velocity dispersion found in several other works (e.g. Cheung et al 2012, Wake et al 2012, Fang et al 2013.…”

“…Baldry et al 2006, Peng et al 2010, Woo et al 2013). However, our result does agree with a complementary analysis, based on the area of the passive fraction relationships, presented in Bluck et al (2015). Furthermore, the finding by Bell et al (2008Bell et al ( , 2012 that essentially all truly passive systems have a high Sérsic index bulge (see also Wuyts et al 2011) is in qualitative agreement with our finding that a high central velocity dispersion and hence central density is the best predictor that a galaxy will be quenched (out of our chosen list of physical galaxy parameters).…”

“…Since our purpose in this paper is to probe galaxy quenching we must carefully correct for this new bias before continuing. For this sub-sample we weight each galaxy in the ANN code by the inverse of the probability of its inclusion (which is a function of its structure, B/T), specifically we calculate (as in Bluck et al 2015):…”

“…For the main sample we use a redshift cut of z spec < 0.2 (as in Bluck et al 2014Bluck et al , 2015. Here we consider restricting the sample…”

Quiescence Correlates Strongly With Directly Measured Black Hole Mass in Central Galaxies

“…The motivation for exploring the relationship between M * , M BH , and quiescence was to test the importance of M BH in driving quiescence. Previous works have linked quiescence with quantities that correlate with M BH , such as velocity dispersion (σ, e.g., Franx et al 2008) or bulge mass (M bul , e.g., Bluck et al 2014a). As such, whether σ or M bul correlates better with quiescence than M BH may provide important physical insight.…”