The Astrophysical Journal, 873:85 (26pp), 2019 March 1 https://doi.org/10.3847/1538-4357/aaf3b8 © 2019. The American Astronomical Society. All rights reserved. Black Hole Mass Scaling Relations for Spiral Galaxies. I. MBH–M*,sph Benjamin L. Davis1 , Alister W. Graham1 , and Ewan Cameron2 1 Centre for Astrophysics and Supercomputing, Swinburne University of Technology, Hawthorn, VIC 3122, Australia; [email protected] 2 Oxford Big Data Institute, University of Oxford, Oxford OX3 7LF, UK Received 2018 May 18; revised 2018 October 19; accepted 2018 November 13; published 2019 March 6 Abstract ( )–( ) The supermassive black hole mass, MBH bulge stellar mass, M*,sph relation is, obviously, derived using two quantities. We endeavor to provide accurate values for the latter via detailed multicomponent galaxy decompositions for the current full sample of 43 spiral galaxies having directly measured MBH values; 35 of these galaxies have been alleged to contain pseudobulges, 21 have water maser measurements, and three appear bulgeless. This more than doubles the previous sample size of spiral galaxies with a finessed image analysis. We have analyzed near-infrared images, accounting for not only the bulge, disk (exponential, truncated, or inclined), and bar but also spiral arms and rings and additional central components (active galactic nuclei (AGNs),etc.). A symmetric Bayesian analysis finds = +0.35 / 10 +( ± ) υ log (MMBH ☉) (2.44-0.31) log{M*,sph [u(1.15´ 10 M☉ )]} 7.24 0.12 ,with a stellar mass-to-light ratio term. The level of scatter equals that about the MBH–σ* relation. The nonlinear slope rules out the idea that many mergers, coupled with the central limit theorem, produced this scaling relation, and it corroborates previous observational studies and simulations, which have reported a near-quadratic slope at the low-mass end of the MBH– M ,sph diagram. Furthermore, bulges with AGNs follow this relation; they are not offset by an order of magnitude, * – and models that have invoked AGN feedback to establish a linear MBH M*,sph relation need revisiting. We additionally present an updated MBH–(Sérsic index, nsph) relation for spiral galaxy bulges with a comparable level of –( f) scatter and a new M*,sph spiral-arm pitch angle, relation. Key words: black hole physics – galaxies: bulges – galaxies: evolution – galaxies: fundamental parameters – galaxies: spiral – galaxies: structure Supporting material: figure set 1. Introduction Graham (2012) highlighted that the (stellar luminosity, L)– (stellar velocity dispersion, s ) relation for low-luminosity early- The connection between supermassive black hole (SMBH) * type galaxies was inconsistent with the M –L and M –σ* mass, M , and host bulge/spheroid stellar mass, M ,has BH BH BH *,sph relations of the day. With L µ s2 at B-band absolute magnitudes been a topic of discussion and publication for nearly three * M -20.5 (Vega) mag (Davies et al. 1983; see also decades, since Dressler (1989; see also Yee 1992) suggested that a B Graham 2016 and references therein) and the literature reporting linear relation exists. Most of the first generation of observational 4 6 2 3 MBH µ s –s , one must have that MLBH µ –L ,where studies of the log MBH–log M ,sph relationship (Kormendy & * * 1 * MB -20.5 mag. This is much steeper than the MBH∝L Richstone 1995;Franceschinietal.1998; Magorrian et al. 1998; relation that had typically been reported in the literature, and it McLure & Dunlop 2002; Marconi & Hunt 2003;Häring&Rix explained the apparent departures at low magnitudes in many past 2004) reported quasi-linear (i.e., approximately linear) log(black – ( )– ( ) MBH L diagrams e.g.,Sanietal.2011; van den Bosch et al. hole mass log spheroid mass scaling relations based on 2012; McConnell & Ma 2013).Graham(2012) therefore luminosities and masses from predominantly high-mass, early- advocated a broken M –M power-law relation with a ( ) BH *,sph type galaxies, supporting the notion of Dressler 1989 .This near-linear slope at the high-mass end and a near-quadratic slope period also introduced the correlation between black hole mass (or slope of ≈2.5) at the low-mass end. σ ( and the spheroid stellar velocity dispersion, * Ferrarese & As discussed in a recent review article by Graham (2016) ) Merritt 2000;Gebhardtetal.2000; Merritt 2000 . regarding the various bulge–(black hole) scaling relations, the ( The second generation of studies e.g., Ferrarese & consequences of this steeper relation are numerous and far- Ford 2005; Graham 2007; Lauer et al. 2007; Gültekin et al. reaching. It has implications ranging from theories of galaxy– 2009; Sani et al. 2011; Beifiori et al. 2012; Erwin & (black hole) coevolution (Graham & Scott 2013) to the design Gadotti 2012; van den Bosch et al. 2012; Vika et al. 2012; of space-based gravitational-wave detectors (Mapelli et al. McConnell & Ma 2013) continued to recover a near-linear 2012). Graham (2012), Scott et al. (2013), and Graham & Scott M –M relation. However, the inclusion of more low-mass ( ) – BH *,sph 2013, 2015 offered an interpretation for the bent MBH M*,sph spheroids revealed departures and produced a notably steeper relation involving core-Sérsic (Graham et al. 2003b) and Sérsic distribution than was seen at the high-mass end. Laor (1998, spheroids at the high- and low-mass ends of the diagram, 2001), Wandel (1999), and Ryan et al. (2007) were the first to respectively. An alternative proposal was offered in Savorgnan realize this, and their work suggested a steeper (single) power et al. (2016) in terms of a red (early-type galaxy) and blue (late- law with a slope of 1.53±0.14, presented in Laor (2001). type galaxy) sequence. Salucci et al. (2000) reported that the M –M relation – BH *,sph While most studies of the MBH M*,sph scaling relation have might be significantly steeper for spiral galaxies than for been dominated by early-type galaxies with high masses 7 (massive) elliptical galaxies based on their study of black holes (MMBH 10 ☉), here we will focus on spiral galaxies. We with upper-limit estimates for their masses. endeavor to provide further insight into the true slope of the 1 The Astrophysical Journal, 873:85 (26pp), 2019 March 1 Davis, Graham, & Cameron – –σ – low-mass end of the MBH M*,sph relation. We do this, in part, calibrated to the MBH * or MBH M*,sph relations. Literature by performing a detailed but justifiable decomposition of the searches on these criteria yield a total of 44 galaxies that have galaxy light into its constituent structural components. Our been classified as spiral types (including one ambiguously work builds upon the study of Savorgnan & Graham (2016a), classified spiral: Cygnus A). Our sample of 44 is ultimately which included 18 spiral galaxies3 with eight galaxies culled down by removing four galaxies: Cygnus A (probably 7 possessing SMBHs less massive than »10 M☉. Using an an early-type galaxy) and three bulgeless galaxies. This yields array of different regression methods, their data did not enable our primary sample of 40 spiral galaxies with bulges. All black them to discriminate between slopes of 2 or 3 for the spiral hole masses have been adjusted from their originally published – fi galaxy MBH M*,sph relation. We have more than doubled the masses to conform with our ducial distances. Correspond- sample size, increasing it to 40 spiral galaxies with bulges, ingly, the error bars on the SMBH masses have been adjusted 7 including 17 galaxies having MMBH < 10 ☉. This will enable considering the uncertainty in the distance to their host – us to better constrain the MBH M*,sph relation for spiral galaxies. All references for our adopted distances and black galaxies, thereby contributing to the discussion as to whether hole masses are compiled in Davis et al. (2017). early- and late-type galaxies follow a different relation, possibly indicative of a different formation channel. 2.2. Imaging Data In addition to the care that has gone into acquiring the bulge ( ) masses that are tabulated here, we have applied a sophisticated Following Sani et al. 2011 and Savorgnan & Graham ( ) μ Bayesian linear regression to obtain the optimal M –M 2016a , we have used 3.6 m imaging from the Spitzer Space BH *,sph / / fi relation. We have also included the results when using the more Telescope for our bulge disk etc. light pro le decompositions fi familiar BCES regression from Akritas & Bershady (1996) and due to the lack of signi cant dust extinction at this 4 μ the modified FITEXY routine (Press et al. 1992; Tremaine et al. wavelength. The 3.6 m imaging data are obtained primarily ( 4 2002). We also briefly present the M –n relation from a from the Spitzer Survey of Stellar Structure in Galaxies SG; BH maj ) 5 sample of only spiral galaxy bulges. Furthermore, given the Sheth et al. 2010 , which provides large mosaicked, processed strong relation between the black hole mass and the disk’s images and masks for all of their galaxies. For galaxies not part 4 spiral-arm pitch angle, f (Seigar et al. 2008; Berrier et al. 2013; of the SG, we obtained processed images from the Spitzer ( ) 6 Davis et al. 2017), we have investigated the relation between Heritage Archive SHA . When both of these sources failed to pitch angle and bulge stellar mass. include imaging of a target galaxy, or the resolution was not In what follows, we first introduce our sample of spiral galaxies adequate to quantify small bulges, alternative processed images ( )7 and their image sources. We then discuss our surface brightness were collected from the Hubble Space Telescope HST or the ( ) 8 profile decomposition technique and the conversion of luminosity Two Micron All Sky Survey 2MASS Large Galaxy Atlas fi ( ) fi into stellar mass (see Section 2).