SUPER-MASSIVE SCALING RELATIONS AND PECULIAR RINGED

A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVERSITY OF MINNESOTA BY

BURCIN MUTLU

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

MARC S. SEIGAR

June, 2017 c BURCIN MUTLU 2017 ALL RIGHTS RESERVED Acknowledgements

There are several people who I would like to acknowledge for directly or indirectly contributing to this dissertation. First and foremost, I would like to acknowledge the guidance and support of my ad- visor, Marc S. Seigar. I am thankful to him for his continuous encouragement, patience, and kindness. I appreciate all his contributions of knowledge, expertise, and time, which were invaluable to my success in graduate school. He has set an example of excellence as a researcher, mentor, and role model. In addition, I would like to thank my dissertation committee, Liliya L. R. Williams, M. Claudia Scarlata, and Robert Lysak, for their insightful input, constructive criticism and direction during the course of this dissertation. I have crossed paths with many collaborators who have influenced and enhanced my research. Patrick Treuthardt has been a collaborator for most of the work during my dissertation. The addition of his scientific point of view has improved the quality of the work in this dissertation tremendously. Our discussions have always been stimulating and rewarding. I am thankful to him for mentoring me and being a dear friend to me. I would also like to thank Benjamin L. Davis for numerous helpful advice and inspiring discussions. He has directly involved with many aspects of Chapter 1. I am also grateful to Joel C. Berrier for his valuable feedbacks on Chapter 2. I would like to particularly acknowledge Alister W. Graham for the ideas that he shared with me and

i for the research directions that he advised me. I am thankful to him for introducing me to the collaboration on the search of intermediate-mass black holes. Within this project, I thank Filippos Koliopanos, Bogdan C. Ciambur, Natalie A. Webb, Mickael Coriat, Olivier Godet, and Didier Barret, and I am looking forward to continue this fruitful collaboration. In addition, I would like to acknowledge Ian B. Hewitt, who contributed to updating the software for the measurements of spiral arm pitch angle, and Lauren E. Koval, who contributed to writing the initial scripts to engage with Illustris data in Chapter 2. I would also like to thank Mithila Mangedarage for the innumerable scientific conversations. In addition, I acknowledge Noah Brosch for his valuable comments and suggestions which have considerably contributed to Chapter 5. In this context, I would like to especially mention the contributions of the referees of the various journals. This research has made extensive use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Tech- nology, under contract with National Aeronautics and Space Administration (NASA). Part of this work is based on observations made with the NASA/ESA (HST), and obtained from the Hubble Legacy Archive, which is a collabo- ration between the Space Telescope Science Institute (STScI/NASA), the Space Tele- scope European Coordinating Facility (ST-ECF/ESA) and the Canadian Astronomy Data Centre (CADC/NRC/CSA). The Mikulski Archive for Space Telescopes (MAST) public access site is also used for retrieving Evolution Explorer (GALEX) Re- lease 7 data products. GALEX is operated for NASA by the California Institute of Technology. I also made use of the data products from the Two Micron All Sky Survey (2MASS), which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by NASA and National Science Foundation (NSF) and data products from the Wide-field Infrared

ii Survey Explorer (WISE), which is a joint project of the University of California, Los An- geles, and the Jet Propulsion Laboratory/California Institute of Technology, funded by NASA. Chapter 3 has made use of the Illustris database. The Illustris project acknowl- edges support from many sources: support by the DFG Research Centre SFB-881 The System through project A1, and by the European Research Council under ERC-StG EXAGAL-308037, support from the HST grants program, number HST-AR- 12856.01-A, support for program #12856 by NASA through a grant from STScI, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555, support from NASA grant NNX12AC67G and NSF grant AST-1312095, support from the Alexander von Humboldt Foundation, NSF grant AST- 0907969, support from XSEDE grant AST-130032, which is supported by NSF grant number OCI-1053575. The Illustris simulation was run on the CURIE supercomputer at CEA/France as part of PRACE project RA0844, and the SuperMUC computer at the Leibniz Computing Centre, Germany, as part of project pr85je. Further simulations were run on the Harvard Odyssey and CfA/ITC clusters, the Ranger and Stampede supercomputers at the Texas Advanced Computing Center through XSEDE, and the Kraken supercomputer at Oak Ridge National Laboratory through XSEDE. I had the privilege of presenting the portions of this dissertation at several confer- ences and seminar talks. On many of these conferences, I met several distinguished researchers with whom I had stimulating discussions. I acknowledge the support of the American Astronomical Society and NSF in the form of an International Travel Grant, which enabled me to attend these conferences, and the generous support of the University of Minnesota Duluth and the Fund for Astrophysical Research. I would also like to acknowledge Adrian Smith who made a short film to present my research on peculiar ringed galaxies to a non-scientific audience and helped me to reach out to the general public. I am grateful to Mahdi Sanati and Beth Thacker for

iii all the support and encouragement they provided throughout my graduate life. I very much appreciate all the assistance Terry J. Jones, Terry Thibeault and Jeanne Peterson provided during my academic stay in Minnesota. I would like to thank the members of the Department of Physics & Astronomy at University of Minnesota, both Twin Cities and Duluth campuses. The faculty, staff, and students really made my life in Minnesota one I will always remember. While graduate school was sometimes stressful, I had close friends whom I could always count on to get me through the tough times. I especially thank Sabriye C¸akmak, Neslihan Bi¸sgin, Cholpon Tilegenova, Hilal Cansızo˘glu and Elife Do˘gan for providing me a lot of support and encouragement over the years. I extend my sincerest thanks to Colleen Seigar, who invited me over every Thanksgiving and created an unique family environment for me. I am eternally grateful to all of my family for supporting me in every way through this journey. First and foremost, I would like to thank my husband, Muhammet Pakdil, for always being there for me, believing in me, and keeping me sane. Without his ongoing support, I could not have finished my research. I am extremely grateful to my father, Baattin Mutlu, and my mother, G¨ulten Mutlu, who raised me and taught me to study hard and to give priority in my life to the quest for knowledge. I owe my deepest gratitude to my sister, Burcu Mutlu, who helped spark and nurture my interest in science. She always inspires me to explore new directions in life and follow my dreams.

iv Dedication

To my mom and dad. This journey would not have been possible if not for them, and I dedicate this dissertation to them.

v Abstract

This dissertation aims to improve the theory of galaxy formation through two indepen- dent areas of investigation: 1) super-massive black hole (BH) scaling relations and 2) formation mechanisms of peculiar rings. Several scaling relations between BH masses and numerous properties of their host galaxies have been reported in the literature, implying a co-evolution scenario of galaxies and their central BHs. The first part of this dissertation explores these scaling relations in both observations and simulations. Chapter 2 presents two important applications of the scaling relations: a determination of black hole mass function (BHMF) and a search for intermediate-mass black holes (IMBHs). I estimated a local BHMF through imaging data only by using the statistically tightest correlations. This work provided a reliable census of local BHs, especially for the low-mass regime. This chapter then focuses on my contributions to a collaborative project in the search for IMBHs, in which I provided the BH mass estimations from the spiral arm morphology. This collaboration demonstrated, for the first time, the consistency between the predictions of several popular scaling relations in the low-mass regime. Chapter 3 explores the BH galaxy − connection beyond the bulge by using the Illustris simulation. This work showed Illustris establishes very tight correlations between the BH mass and large-scale properties of the host galaxy, not only for early-type galaxies but also late-type galaxies, regardless of bar morphology. These tight relations suggest that halo properties play an important role in determining those of the galaxy and its BH. The main focus of Chapters 4 and 5 is ring formation mechanisms, in particular the origin of Hoag-type galaxies. Studying such peculiar galaxies is important to address how different kinds of interactions contribute to different galaxy morphologies. Chapter

vi 5 presents a photometric study of PGC 1000714, a galaxy with a fair resemblance to Hoag’s Object. This work has revealed, for the first time, an with two fairly round rings. A number of formation scenarios are discussed. However, there are many questions yet to address, especially the origin of the inner ring. This dissertation is just a beginning to understanding this interesting group of galaxies.

vii Contents

Acknowledgements i

Dedication v

Abstract vi

List of Tables xi

List of Figures xii

1 Introduction 1

1.1 Do black holes exist in all sizes? ...... 2 1.1.1 TheSearchforIMBHs...... 3 1.2 BHScalingRelations...... 5 1.2.1 Connection with Bulge Properties ...... 5 1.2.2 BH Scaling Relations Beyond the Bulge ...... 6 1.3 TheOriginofGalacticRings ...... 8 1.4 Outline ...... 9

2 Testing Black Hole Scaling Relations Using Observational Data 11

2.1 Overview ...... 12

viii 2.2 The Local Black Hole Mass Function ...... 14 2.2.1 The M P and M n Relations ...... 14 BH − BH − 2.2.2 Data and Sample Selection ...... 16 2.2.3 S´ersicIndexMeasurement ...... 23 2.2.4 S´ersicIndexDistribution ...... 34 2.2.5 EstimatingBHMF ...... 36 2.2.6 BHMassDensity...... 40 2.2.7 Discussion...... 40 2.3 TheSearchforIMBHs ...... 50 2.3.1 Methodology ...... 50 2.3.2 SpiralArmsAnalysis...... 51 2.3.3 Results ...... 54 2.4 Conclusions ...... 57

3 The Super-massive Black Hole Galaxy Connection in Illustris 60 − 3.1 Overview ...... 60 3.2 The Illustris Spiral Arm Morphology and Pitch Angle Scaling Relations 62 3.2.1 Spiral Arm Morphology in Multi-wavebands ...... 62 3.2.2 Pitch Angle Scaling Relations ...... 69 3.3 Black Hole Scaling Relations Beyond the Bulge ...... 73 3.3.1 The M M Relation ...... 74 BH − ⋆,total 3.3.2 The M M Relation ...... 75 BH − DM 3.3.3 The M M Relation ...... 75 BH − halo 3.4 Conclusions ...... 81

4 The Ring Formation Mechanisms 86

4.1 ResonanceRings ...... 87

ix 4.2 Galaxy GalaxyInteractions...... 89 − 4.3 Hoag-typeGalaxies...... 92

5 A photometric study of the peculiar ringed galaxy: PGC 1000714 96

5.1 Overview ...... 96 5.2 ObservationalData...... 98 5.3 GeneralProperties ...... 101 5.4 IsophotalAnalysis ...... 105 5.4.1 EllipseFitting ...... 105 5.4.2 RadialProfile...... 109 5.5 2DImageAnalysis ...... 109 5.6 Rings ...... 117 5.7 StellarPopulations ...... 119 5.8 Discussion...... 121 5.9 Conclusions ...... 128

Appendix A. Acronyms 163

Appendix B. News Release on PGC 1000714 173

x List of Tables

2.1 Early-typesampleforlocalBHMF...... 31 2.2 BHMFvalues...... 47 2.3 BHMFevaluation...... 48 2.4 Multiplemethodapproach...... 56 3.1 Illustrisspiralgalaxysample ...... 82 5.1 General information for PGC 1000714 from the NED ...... 104 5.2 Structural parameters of PGC 1000714 obtained by GALFIT ...... 115 5.3 The observed SED of the core and the outer ring of PGC 1000714 . . . 120 5.4 Comparison of PGC 1000714 with other galaxies ...... 127 A.1 Acronyms ...... 163

xi List of Figures

1.1 Black hole mass versus pitch angle relation by Berrier et al. (2013) . . . 7 2.1 Volume-limited sample selection...... 18 2.2 Completeness check-1: Luminosity function ...... 20 2.3 Completeness check-2: Distribution of morphological types ...... 21 2.4 Completeness check-3: Distribution of T-types ...... 22 2.5 The methodology for estimating the errors on the model parameters . . 24 2.6 The surface brightness profiles for 30 elliptical galaxies...... 25 2.7 The surface brightness profiles for 38 S0 galaxies ...... 27 2.8 TheS´ersicindexhistogram ...... 35 2.9 Early-typeBHMF ...... 37 2.10All-typeBHMF...... 39 2.11 Comparison of the early-type BHMF with previous works ...... 42 2.12 Comparison of the all-type BHMF with previous works ...... 44 2.13 Comparison of the all-type BHMF with more recent works ...... 45 2.14 An example measurement of the pitch angle ...... 52 2.15 Mass estimations from multiple method approach ...... 55 3.1 Simulated spiral galaxies in Illustris ...... 64 3.2 An example measurement of the pitch angle in Illustris ...... 66 3.3 Illustris spiral arm morphology in multi-wavebands ...... 68

xii 3.4 The M P relationinIllustris ...... 71 BH − 3.5 The M P and M P relations in Illustris ...... 72 DM − halo − 3.6 The M M relation in Illustris ...... 76 BH − ⋆,total 3.7 The M M relationinIllustris ...... 77 BH − DM 3.8 The M M relationinIllustris ...... 78 BH − halo 4.1 Typesofresonancerings...... 87 4.2 Collisional ring galaxy: the Carthwheel Galaxy ...... 90 4.3 Accretion ring (NGC 7742) and PRG (NGC 4650A) ...... 91 4.4 Hoag’sObject...... 94 5.1 Sky estimation method used for PGC 1000714 ...... 99 5.2 The images of PGC 1000714 in multi-wavebands ...... 102 5.3 The high frequency residual image for PGC 1000714 ...... 102 5.4 The residual images of PGC 1000714 (FMEDIAN and BMODEL) . . . 103 5.5 The geometric properties of PGC 1000714 ...... 108 5.6 The color gradients of PGC 1000714 ...... 110 5.7 The best S´ersic model fit of the core of PGC 1000714 in the BVRI -bands 111 5.8 The 2D B I color index map of PGC 1000714 ...... 112 − 5.9 The 2D image decomposition of PGC 1000714 by GALFIT ...... 113 5.10 The light profiles of PGC 1000714 with the outer ring ...... 118 5.11 The observed SED of the core and the outer ring of PGC 1000714 . . . 121

xiii Chapter 1

Introduction

Galaxies are fundamental units of matter in the universe, and an understanding of galaxy formation and evolution is one of the major goals of extragalactic studies. This dissertation focuses on two important aspects of galaxy evolution: 1) the link between galaxies and their super-massive black holes and 2) the origin of galactic rings. It is now well accepted that super-massive black holes (BH) reside at the center of most massive galaxies, both quiescent and active (Kormendy and Richstone 1995; Barth 2004; Kormendy 2004; Magorrian et al. 1998). In the last few decades, several scaling relationships between the mass of the central BH and the overall morphology and dynamics of the host galaxy have been found and this points to a co-evolution scenario of galaxy formation and BH growth. A common interpretation for these relationships is that BHs regulate their own growth and that of their host galaxies through AGN- feedback (e.g., Silk and Rees 1998; Springel et al. 2005; Bower et al. 2006; Croton et al. 2006; Di Matteo et al. 2008; Ciotti et al. 2009; Fanidakis et al. 2011). Therefore, these scaling relations have important implications not only for BHs and galaxies but also understanding the effects of AGN feedback. In addition, a well-defined relation provides an easy and accurate means to estimate BH masses in large samples of galaxies, or in

1 2 individual objects with insufficient data to measure BH masses from the dynamics of , gas, or masers. The central theme of Chapters 2 and 3 is the scaling relations between BHs and their host galaxies. Another interesting problem in the studies of galaxy evolution is the origin of galac- tic rings. Up to 20-30% of all spiral disk galaxies show a ring-shaped pattern in their light distribution (Buta and Combes 1996). They are closely connected with such fun- damental aspects as the distribution of the dark matter in galaxies, internal secular evolution of galactic substructures and environmental interactions (e.g., Whitmore et al. 1987, 1990; Sackett et al. 1994; Combes and Arnaboldi 1996; Buta and Combes 1996; Schweizer et al. 1987; Bekki 1997, 1998; Bournaud and Combes 2003; Macci`oet al. 2006; Brook et al. 2008). An especially interesting ring case is Hoag’s Object (PGC 054559; Hoag 1950) with its peculiar morphology: an elliptical-like core with a nearly perfect outer ring, and no signs of bar or stellar disk. Hoag-type galaxies, which bear strong resemblance to Hoag’s Object, are extremely rare (Schweizer et al. 1987) and their origin is still debated. The central theme of Chapters 4 and 5 is the ring formation mechanisms, with a particular interest and focus on the origin of Hoag-type galaxies. This first chapter gives a basic overview, briefly introduces BH scaling relations as well as the ring formation mechanisms, and finally discusses the research goals.

1.1 Do black holes exist in all sizes?

Super-massive black holes (BHs) contain between a few million and a few billion times the mass of the Sun and reside at the the centers of most galaxies, including the cen- ter of our own galaxy, the Milky Way (e.g., Sch¨odel et al. 2002). Their origin is still unknown. At the other extreme, stellar-mass black holes (s-BHs) have masses of be- tween a few times and several tens of times that of the Sun and are end-products of stellar evolution. While the existence of BHs and s-BHs has been firmly established, 3 there is a notable scarcity of black holes in the intermediate mass range, i.e., the so- called intermediate-mass black holes (IMBHs). Although their existence has not been established with certainity yet, they are predicted by a variety of plausible scenarios (evolution of Population III stars: e.g., Madau and Rees 2001; Schneider et al. 2002; Ryu et al. 2016, repeated mergers onto massive stars in young massive clusters: e.g., Portegies Zwart and McMillan 2002; Portegies Zwart et al. 2004; G¨urkan et al. 2004; Freitag et al. 2006, or black hole mergers in dense clusters: e.g., Quinlan and Shapiro 1990; Lee 1995; Mouri and Taniguchi 2002; Miller and Hamilton 2002a,b) and there are observational hints for IMBHs (e.g., Gebhardt et al. 2005; Gerssen et al. 2002; Filippenko and Ho 2003; Trenti et al. 2007; Ulvestad et al. 2007, cf. Baumgardt et al. 2003 for an alternative view).

1.1.1 The Search for IMBHs

The search for IMBHs is a young and promising field of research. IMBHs have been invoked to explain the origin of the ultra-luminous X-ray sources (ULXs, e.g., Foschini et al. 2002; Zezas and Fabbiano 2002). However, it was later shown that most ULXs could be explained by super-Eddington accretion onto quite massive s-BHs (e.g., Sutton et al. 2013; Bachetti et al. 2013). So far, only one small group of hyper-luminous X-ray sources (HLXs), in particular ESO 243-49 HLX-1, has been reliably identified as an accreting IMBH candidate (e.g., Farrell et al. 2009), which cannot be explained by invoking super-Eddington accretion and/or beaming of the X-ray emission. Currently, many research groups are pursuing the search of other IMBH candidates similar to ESO 243-49 HLX-1. In addition to the search for HLXs, many attempts have been made to detect IMBHs in globular clusters, ranging from spectroscopic and photometric velocity measurements (e.g., Noyola et al. 2008) to X-ray and radio observations in order to find signatures of 4 accretion (e.g., Strader et al. 2012). The results have been full of ambiguity, unsatisfying upper limits and surprising evidential reveals. For example, while kinematic modelling of the Mayall II indicated that it contains an IMBH (Gebhardt et al. 2005), alternative dynamical modelling has shown that the data can be explained with- out such an object (Baumgardt et al. 2003). Similar modelling attempts have only been able to provide upper limits on the mass of potential central IMBHs (Gerssen et al. 2002; McLaughlin et al. 2006). An IMBH candidate, M82 X-2, was revealed to be a neutron , not to be a black hole at all (Bachetti et al. 2014). However, Kızıltan et al. (2017) have recently reported evidence of a central black hole in 47 Tucanae with a mass of 2,200 solar masses when the dynamical state of the globular cluster is probed with pul- sars. This is the best evidence yet for an IMBH in a globular cluster (G¨ultekin 2017). Although this new method is very powerful for detecting black holes with no electro- magnetic counterpart, it only works for globular clusters with the right combination of distance and bright pulsars (Kızıltan et al. 2017). Moreover, the detection of AGNs in low-luminosity late-type galaxies strongly suggests the existence of IMBHs (Filippenko and Ho 2003; Barth et al. 2004; Greene and Ho 2004), and has motivated the search for IMBHs in low-mass galax- ies. The study of low-mass galaxies using the M σ relation, to be discussed in BH − ⋆ the next section, has led to the identification of multiple (tentative) IMBH candidates (e.g., Greene and Ho 2004; Barth et al. 2008; Dong et al. 2012). However, these samples suffer, to some degree, from luminosity bias and large uncertainties that are partially caused by the precarious application of the M σ relation in the low-mass regime BH − ⋆ (see discussion in Greene and Ho 2007; Moran et al. 2014). It is, therefore, crucial to advance the field of IMBHs in all directions. 5 1.2 BH Scaling Relations

Correlations between black hole masses and numerous properties of their host galaxies have been explored in the literature. This section summarizes BH scaling relations that have been shown to exist with bulge properties, and also beyond the bulge.

1.2.1 Connection with Bulge Properties

For the last few decades, several observational studies have continually showed that BH masses (MBH ) correlate with the bulge component of their host galaxies, indicating that BH and bulge formation are tightly connected. These studies recovered scaling relations between MBH and stellar velocity dispersion (σ⋆) (e.g., Ferrarese and Merritt 2000; Gebhardt et al. 2000; Merritt and Ferrarese 2001; Tremaine et al. 2002; Wyithe 2006; Hu 2008; G¨ultekin et al. 2009; Schulze and Gebhardt 2011; Graham et al. 2011;

Beifiori et al. 2012) and between MBH and the of the bulge (M⋆,Bulge) (e.g., Magorrian et al. 1998; Marconi and Hunt 2003; H¨aring and Rix 2004; Sani et al. 2011; Beifiori et al. 2012; Savorgnan et al. 2016). The velocity dispersion is a measure of the random velocity distribution of stars orbiting in the bulge and is directly related to bulge dynamical mass, i.e., M 3R σ /G (Marconi and Hunt 2003) where R is ⋆,Bulge ∼ e ⋆ e the effective radius of the bulge. The more massive the bulge is, the more massive the BH.

Various scaling relations between MBH and the photometric properties of the bulge have also been reported: bulge optical luminosity (e.g., Kormendy and Richstone 1995; Kormendy and Gebhardt 2001; G¨ultekin et al. 2009; Schulze and Gebhardt 2011;

McConnell et al. 2011; Beifiori et al. 2012), bulge near-infrared luminosity (LBulge) (e.g., Marconi and Hunt 2003; McLure and Dunlop 2002; Graham 2007; Sani et al. 2011), and bulge concentration or the S´ersic index (n) (e.g., Graham et al. 2001; 6 Graham and Driver 2007; Savorgnan 2016). Although many of these correlating prop- erties of the bulge require spectroscopic measurements, one which does not is the S´ersic index (Graham and Driver 2007). It has the advantage that it can be measured from photometrically uncalibrated images alone. The substructures in the most commonly cited BH scaling relations (e.g., M BH − L , M σ , M M ) are reported in such a way that galaxies with Bulge BH − ⋆ BH − ⋆,Bulge different properties (e.g. pseudobulges versus real bulges, barred versus non-barred, or cored versus power-law ellipticals) may correlate differently with their central BH masses (e.g., Graham 2008; Hu 2008; McConnell and Ma 2013; Kormendy and Ho 2013; Gebhardt et al. 2011). Graham (2016) claims a bend in the M L rela- BH − Bulge tion and no bend in the M σ relation whereas Kormendy and Ho (2013) and BH − ⋆ McConnell and Ma (2013) claim no bend in the M L relation and a bend in BH − Bulge the M σ relation. The true nature of galaxy evolution in different galaxy types BH − ⋆ still needs to be resolved.

1.2.2 BH Scaling Relations Beyond the Bulge

On a larger scale, various scaling relations beyond the bulge have also been re- ported: total luminosity (e.g., Kormendy and Gebhardt 2001; Beifiori et al. 2012) and total stellar mass (M⋆,total) (e.g., Reines and Volonteri 2015). Most self-regulating theoretical models of BH formation predict a fundamental connection between

MBH and the total gravitational mass of the host galaxy (e.g., Adams et al. 2001; Monaco et al. 2000; Haehnelt et al. 1998; Silk and Rees 1998; Haehnelt and Kauffmann 2000; Cattaneo et al. 1999; Loeb and Rasio 1994; El-Zant et al. 2003). The galaxy mod- els, which study the interaction between the dark matter halos of galaxies and baryonic matter, predict that halo properties determine the masses of the bulge and central BH 7

Figure 1.1: Black hole mass versus pitch angle for spiral galaxies with directly mea- sured black hole masses. Black hole masses measured using stellar and gas dy- namics techniques are depicted by black crosses (10 points), reverberation mapping masses by red triangles (12 points), and maser measurements by blue squares (12 points). The black solid line shows the best linear fit to the data: log(MBH /M⊙) = (8.21 0.16)(0.062 0.009) P , with a reduced χ2 = 4.68 and a scatter of 0.38 dex. This ± ± | | figure is reproduced from Berrier et al. (2013). 8 (Cattaneo 2001; El-Zant et al. 2003; Hopkins et al. 2005). However, there are inconsis- tencies in the observational studies: correlations between MBH and the circular veloc- ity (vc) or the mass (Mhalo) have been reported (e.g., Ferrarese 2002; Baes et al. 2003; Seigar 2011; Volonteri et al. 2011) as well as disputed (e.g., Zasov et al. 2005; Kormendy and Bender 2011; Beifiori et al. 2012). It is not an easy task to inves- tigate this relation observationally due to the uncertainities in measurement of the halo mass. While it only operates in spiral galaxies, a tight correlation between BH mass and spiral arm pitch angle (P ) has been first demonstrated by Seigar et al. (2008), then con- firmed by Berrier et al. (2013). The pitch angle can be described as the angle between the line tangent to a circle and a line tangent to a logarithmic spiral. Small pitch angles, i.e., tightly wound spirals, are associated with large BH masses while high pitch angles, i.e., loosely wound spiral arms, are associated with small BH masses (see Figure 1.1).

1.3 The Origin of Galactic Rings

A significant fraction of galactic rings are formed in the disks as a result of internal dynamics. The main interpretation has been related to orbital resonances in bars or other common non-axisymmetric perturbations, such as ovals, in the galactic gravita- tional potential (Buta and Combes 1996). Although much rarer, there are also peculiar rings that are formed as a result of environmental interactions. The main classes of these peculiar rings are collisional rings (e.g., the Cartwheel Galaxy), polar rings (e.g., A 0136-0801, Whitmore et al. 1990) and accretion rings (e.g., IC 2006, Schweizer et al. 1989). While collisional rings are thought to be products of a strong head-on collision with a companion (Arp 1966; Appleton and Struck-Marcell 1996), accretion rings are thought to be results of gas accretion (Schweizer et al. 1987, 1989). Polar rings are cases where the ring is nearly orthogonal to the plane of the main galaxy, which is often an 9 S0 galaxy (Whitmore et al. 1990). They are believed to be formed due to the galactic merger or accretion of the matter from a nearby system or intergalactic medium (IGM) onto the main galaxy (Bournaud and Combes 2003; Combes 2006; Macci`oet al. 2006).

1.4 Outline

An important step towards understanding BH scaling relations is to estimate demo- graphic information (i.e., masses) for the population of BHs in our universe by studying a complete sample of observable galaxies. Chapter 2 initially uses the statistically tigh- est relations (the M n relation for early-type galaxies and the M P relation for BH − BH − spiral galaxies) and provides reliable data on local black hole mass function (BHMF) through imaging data alone. A robust BHMF is important because it helps to de- scribe the evolution of the BH distribution and provides important constraints on the co-evolution of the quasar and black hole populations, thus improving our understand- ing of the accretion history of the universe. This chapter then focuses on my contri- butions to a collaborative project in the search for IMBHs in nearby low-luminosity AGN (LLAGN) and dwarf galaxies, which have been reported to have candidate central IMBHs by Graham and Scott (2013). The search for IMBHs is important, not only because their existence is predicted by plausible scenarios, but also because they are proposed to be the seeds for BHs. Discovering IMBHs and studying their observational signatures will elucidate the mechanisms of early galaxy formation. An additional and equally important aim of this collaboration is to investigate the consistency between the predictions of several popular BH scaling relations and the fundamental plane of BH activity, i.e., X-ray and radio luminosity relations in accreting BHs. It is important to study the properties of BHs and their host galaxies through both observations and simulations, and compare their results in order to understand their physics and formative histories. Chapter 3 explores the BH-galaxy connection beyond 10 the bulge in Illustris, which is a large-scale and high-resolution cosmological simulation. Chapter 4 reviews possible scenarios of ring formation, focusing on the aspects most relevant to the rest of this work: the origin of Hoag-type galaxies. This review is necessarily brief, and more information on the related subjects can be found in the references mentioned. Hoag-type galaxies are extremely rare and peculiar systems, therefore it is necessary to increase the sample of known objects by performing detailed studies on other possible candidates to derive conclusions about their nature, evolution, and systematic proper- ties. Chapter 5 presents a photometric study of PGC 1000714, a galaxy resembling Hoag’s Object with a complete detached outer ring, that has not yet been described in the literature. Its fair resemblance to Hoag’s Object makes it a good target for a detailed study of Hoag-type galaxy structure. Such peculiar systems help our under- standing of galaxy formation in general, since they represent extreme cases, providing clues on formation mechanisms. Chapter 2

Testing Black Hole Scaling Relations Using Observational Data

A slightly modified version of Section 2.2 has been published in Astrophysical Journal with the following bibliographic reference: B. Mutlu-Pakdil, M. S. Seigar, and B. L. Davis. The Local Black Hole Mass Function Derived from the M P and the M n BH − BH − Relations. ApJ, 830:117, October 2016. A modified version of Section 2.3 has been published in Astronomy & Astrophysics Journal with the following bibliographic reference: F. Koliopanos, B. C. Ciambur, A. W. Graham, N. A. Webb, M. Coriat, B. Mutlu-Pakdil, B. L. Davis, O. Godet, D. Barret, and M. S. Seigar. Searching for intermediate-mass black holes in galaxies with low-luminosity AGN: a multiple-method approach. A&A, 601:A20, April 2017.

11 12 2.1 Overview

This chapter presents two important applications of BH scaling relations: a determina- tion of the mass function of the central BHs (BHMF) in the local universe (Section 2.2) and a search for intermediate-mass black holes (IMBHs) (Section 2.3). A robust BHMF helps to describe the evolution of the BH distribution and provides important constraints on the co-evolution of the quasar and black hole populations. The most well-known theoretical constrains are on the integrated emissivity of the quasar population, integrated mass density of black holes, and the average black hole accretion rate (e.g., Soltan 1982; Fabian and Iwasawa 1999; Elvis et al. 2002; Shankar et al. 2009). A comparison among the recent local BHMF estimates derived from different scaling relations can be seen in Figure 5 of Shankar et al. (2009). Most of these studies use an analytical approach, which combines measurements of the galaxy luminosity or velocity function with one of BH scaling relations as outlined by H¨aring and Rix (2004). These studies use some assumptions of the morphological type fractions and the bulge-to- total luminosity (B/T ) ratios. The sensitivity of the low-mass end of the BHMF based on these assumptions is well presented in Figure A2 of Vika et al. (2009). Recently, Davis et al. (2014) estimated a BHMF by using the BH mass versus spiral arm pitch angle relation (M P ) for a nearly complete sample of local spiral galaxies in order to BH − produce reliable data for the low-mass end of the local BHMF. In Section 2.2, the aim is to estimate a local BHMF for all galaxy types within the same volume limits in order to complement this late-type BHMF. Therefore, the identical sample selection criteria in

Davis et al. (2014) is used: limiting luminosity (-independent) distance, DL = 25.4 Mpc, and a limiting absolute B-band magnitude, M = 19.12. These limits B − define a sample of 140 spiral, 30 elliptical (E), and 38 lenticular (S0) galaxies. I establish the S´ersic index distribution for early-type (E/S0) galaxies in this sample. Davis et al. (2014) established the pitch angle distribution for their sample, which is identical to 13 the late-type (spiral) galaxy sample in this work. The pitch angle and the S´ersic index distributions are then used to estimate the BHMF for the total volume-limited sample. The observational simplicity of this approach relies on the empirical relation between the mass of the central BH and the S´ersic index for an early-type galaxy or the logarithmic spiral arm pitch angle for a . A cosmological model with ΩΛ = 0.691, ΩM =

−1 −1 0.307, ωb = 0.022 and h67.77 = Ho/(67.77 km s Mpc ) (Planck Collaboration et al. 2014) is adopted throughout Section 2.2. Section 2.3 focusses on the search for IMBHs. As they are proposed to be the seeds for BHs, discovering IMBHs is important to further our understanding of when in the early universe BHs first formed and how they have affected the lives of the galaxies in which they reside. In principle, IMBHs can exist in the centers of low-luminosity AGN (LLAGN) galaxies because these low-mass galaxies are expected to have quiet merger histories, and, therefore, they are more likely to host lower mass central BHs, some of which are expected to lie in the IMBH range. Recently, Graham and Scott (2013) reported a list of 40 IMBH candidates in LLAGN and dwarf galaxies. This list motivated a collaborative project in search of IMBHs in nearby LLAGN galaxies, which have candidate central IMBHs. Section 2.3 reports the first results of this endeavor, which target seven LLAGN galaxies in that list. The most up-to-date BH scaling relations and the correlation between the radio luminosity, X-ray luminosity and the BH mass, known as the fundamental plane of black-hole activity (F P BH), are used to independently − estimate the central black hole mass in all seven galaxies. My contribution was to provide the mass estimations from the spiral arm morphology by using the optical observations (Section 2.3.2). The main results of this work are summarized in Section 2.3.3. 14 2.2 The Local Black Hole Mass Function

2.2.1 The M P and M n Relations BH − BH − The M M , M L , and M n relations all depend on the success BH − ⋆,Bulge BH − Bulge BH − of the measurements of the bulge. In late-type galaxies, there can be difficulties when it comes to isolating the bulge from other components of galaxies (e.g., bars, disk, and spiral arms). In the study of disk galaxies, a standard practice is to assume a fixed value of B/T ratio. This introduces a bias on a BHMF such that BH mass is overestimated in late-type disk galaxies and underestimated in early-type disk galaxies (Graham et al. 2007). Another approach is to use the average B/T ratios derived from R1/n-bulge + exponential-disk decompositions (Graham et al. 2007), which requires heavy image processing tools. The large scatter in these relations to estimate BH mass can be traced back to the complexity of the decomposition in late-type galaxies, particularly in barred galaxies. The M σ relation has had considerable success in estimating BH BH − ⋆ masses in many galaxies. However, it requires spectroscopic measurements, which are observationally expensive and depend on the spectroscopic bandwidth. Furthermore, a careful approach is needed such that a consistent bulge region is always sampled for the measurement of σ⋆. Similar to the above relations, measuring σ⋆ is more complex for disk galaxies than it is for elliptical galaxies because the velocity dispersion from the motion of disks and bars is coupled with σ⋆ and they need to be handled properly (Hu 2008). Among other relations, the M P relation seems promising for late-type galaxies. BH − Berrier et al. (2013) established a linear M P relation for local spiral galaxies as BH − log(M /M⊙) = (8.21 0.16) (0.062 0.009) P with a scatter less than 0.48 dex BH ± − ± | | in all of their samples. This is lower than the intrinsic scatter ( 0.56 dex) of the ≈ M σ relation, using only late types (G¨ultekin et al. 2009). The P -derived BH BH − ⋆ 15 mass estimates also seem to be consistent in galaxies with pseudobulges, where other relations seem to fail (Berrier et al. 2013). Although there are obvious advantages in using the M P relation in late-type galaxies (see Discussion in Berrier et al. 2013), BH − one needs to use a complimentary relation for E/S0 galaxies because the M P BH − relation is only applicable for spiral galaxies. Figure 6 in Berrier et al. (2013) presents evidence that n- and P -derived mass estimates are compatible for non-barred galaxies, and a combination of these two approaches (i.e., using S´ersic index for E/S0 galaxies, and pitch angles for spiral galaxies) may produce a very accurate BHMF for all galaxy types by using only imaging data. Graham et al. (2001) presented evidence that the light concentration of the spheroids correlates well with their BH mass, showing that more centrally concentrated spheroids have more massive black holes. Given that the S´ersic index, n, is essentially a measure- ment of the central light concentration, Graham and Driver (2007) found a log quadratic relation between n and MBH :

n n 2 log(M /M⊙) = (7.98 0.09) + (3.70 0.46) log( ) (3.10 0.84)[log( )] (2.1) BH ± ± 3 − ± 3

+0.07 with an intrinsic scatter of ǫint = 0.18−0.06 dex. Recently, Sani et al. (2011), Vika et al. (2012) and Beifiori et al. (2012) failed to recover a strong M n relation. Savorgnan et al. (2013) re-investigated and recov- BH − ered the relation using a large collection of literature S´ersic index measurements using R-band (Graham and Driver 2007), I-band (Beifiori et al. 2012), K-band (Vika et al. 2012), and 3.6µm (Sani et al. 2011) imaging data. Savorgnan et al. (2013) discussed the systematic effects associated with measuring S´ersic index in different optical and in- frared wavebands. They concluded that the differences expected from measuring S´ersic index in different wavebands are smaller than the differences expected due to other systematic biases such as one-dimensional (1D) decomposition versus two-dimensional 16 (2D) decomposition, or the differences between measuring the S´ersic index along a mi- nor axis versus measuring it along a major axis. Indeed, one migh expect that a S´ersic index measured using a 1D fit (as performed in this chapter) to be 10% smaller than ∼ that measured using a 2D fit (Ferrari et al. 2004). Furthermore, when measuring S´ersic indices in multiple wavebands for the same galaxies, Savorgnan et al. (2013) found that wavelength bias was completely dominated by these other biases, which could be as large as 50%. Given the result of Kelvin et al. (2012), one would expect the S´ersic index measured at 3.6µm to be less than 10% higher than that measured in R-band, which is signifantly smaller than the 50% number given by Savorgnan et al. (2013). Savorgnan et al. (2013) excluded the outlying S´ersic indices, averaged the remaining values, and recovered the M n relation by showing that elliptical and disk galax- BH − ies follow two different linear M n relations. They discussed how this relation is BH − consistent with what would be derived by combining the M L and L n BH − Bulge Bulge − relations and how this explains the log quadratic nature of the M n relation reported BH − by Graham and Driver (2007). In this chapter, early-type galaxies were defined as elliptical and S0 galaxies. The sample used by Graham and Driver (2007) was dominated ( 89%) by elliptical and ∼ S0 galaxies. However, Savorgnan et al. (2013) studied S0 galaxies together with spiral galaxies. Therefore, the log quadratic M n relation reported by Graham and Driver BH − (2007) was used to estimate BH masses in the early-type sample in this work.

2.2.2 Data and Sample Selection

Davis et al. (2014) based their selection criterion on the Carnegie-Irvine Galaxy Sur- vey (CGS) (Ho et al. 2011); it is an almost-complete sample of 605 nearby galaxies in the southern hemisphere. Using the spiral galaxies in this parent sample plus Milky Way, they defined a volume-limited sample which consisted of spiral galaxies within a 17 luminosity (redshift-independent) distance of 25.4 Mpc and a limiting absolute B-band magnitude of M = 19.12. Here, the same selection criterion was followed, except B − also including E and S0 galaxies. As a result, the volume-limited sample consists of 208 host galaxies (30 E, 38 S0, and 140 spiral galaxies) within a comoving volume of V = 3.37 104 h−3 Mpc3 over a lookback time, t 82.1 Myr. Images of selected c × 67.77 L ≤ galaxies were then downloaded from the NASA/IPAC Extragalactic Database (NED). A complete sample selection is necessary to estimate a meaningful BHMF. Therefore, the completeness of the sample was tested within the limits of luminosity distance and absolute B-band magnitude in several ways. First, the sample size was compared with the maximum number of galaxies within these limits. Figure 2.1 shows that the maximum number of galaxies, which is 217, appears at DL = 28.05 Mpc and MB = 19.37, whereas the sample consists of 208 galaxies. While these two limits just differ − by 4%, using the limiting M = 19.12 allowed for inclusion of galaxies with dimmer B − intrinsic brightness, and this helped the sample to be more complete. In addition, the luminosity function was derived in order to check if the volume- limited sample was a fair representation of the local galaxy population over the absolute magnitude range 19.12 . M . 23. The luminosity function was determined as − B − φ(MB ) = ∂N/∂MB , where N is the number of galaxies in the sample in terms of the absolute B-band magnitude and dividing it by the comoving volume of the volume- limited sample. This is illustrated in Figure 2.2, which shows the comparison with the luminosity functions for the overall CGS sample (Ho et al. 2011) and the much larger sample of Blanton et al. (2003) and Bernardi et al. (2013). The luminosity functions of Blanton et al. (2003) and Bernardi et al. (2013) have been shifted by B r = 0.67 − mag, the average color of an Sbc spiral (Fukugita et al. 1995), which is roughly the median Hubble type of both CGS and the volume-limited sample, and also transformed

−1 −1 to H0 = 67.77 km s Mpc . While Blanton et al. (2003) derived the luminosity 18

Figure 2.1: Left: Luminosity distance vs absolute B-band magnitude for all type galaxies (606) found using the magnitude-limiting selection criteria (m 12.9 and δ < 0). B ≤ The upper limit absolute magnitude is modeled as the same exponential in Davis et al. (2014) and is plotted here as the solid black solid line. The dashed red rectangle shows the galaxies in the volume-limited sample. Right: Histograms showing the number of galaxies (spiral galaxies: blue; early-types (E/S0): red; all-types: pink) contained in the limits of luminosity distance and absolute B-band magnitude as the limits are allowed to change on the exponential line based on the limiting luminosity distance. Note that the peak for all galaxy types appears in the histogram with 217 galaxies at DL = 28.05 Mpc. Black dashed line represents DL = 25.4 Mpc that is the limit used in Davis et al. (2014), and gives 208 galaxies (68 E/S0 + 140 spiral), which is very close to the actual peak. Using DL = 25.4 Mpc helps the sample to be complete for dimmer galaxies. Complete volume-limited samples were computed for limiting luminosity distances in increments of 0.001 Mpc. 19 function of z 0.1 galaxies from (SDSS) by using the S´ersic ≈ parameters from a 1D radial surface brightness profile, Bernardi et al. (2013) derived it by using the 2D fits to the whole galaxy image. The overall CGS sample has a luminosity function that agrees quite well with that of Blanton et al. (2003) (Ho et al. 2011). However, the sample has a luminosity function that implies that it was observed in an overdense volume (see red data points in Figure 2.2). Therefore, the luminosity function is renormalized by adding 0.25 on the y-axis in order to be consistent with that − of Blanton et al. (2003) and CGS (see pink data points in Figure 2.2). For the BHMF estimation, the same normalization factor is used (see Section 2.2.5). In addition, due to the sample selection criterion, the luminosity function does not extend below the magnitude limit of M = 19.12. This fact is obviously of interest to the BHMF B − estimation that will be discussed more in Section 2.2.7. Furthermore, the distributions of morphological types of CGS and the sample were compared. The morphological fractions, ftype, of the sample are as such: fE = 0.14, fS0 = 0.18, and fSpiral = 0.67. This is in good agreement with the ones (fE = 0.11 0.03, f = 0.21 0.05, f = 0.62 0.14) reported by Fukugita et al. ± S0 ± Sab+Sbc+Scd ± (1998). Moreover, Figure 2.3 shows that the volume-limited sample preserves the distri- bution of morphological types in CGS. In addition, the T-type distributions of CGS and this sample are compared in Figure 2.4. The T-type values were taken from http://cgs.obs.carnegiescience.edu/CGS/database_tables. The differences be- tween the densities of each T-type are always less than 5%. Imaging data was taken from the NED (see Table 2.1). The absolute magnitudes were calculated from apparent magnitudes, from HyperLeda (Paturel et al. 2003), lumi- nosity distances were compiled from the mean redshift-independent distance from the NED, and extinction factors in the B-band were taken from Schlafly and Finkbeiner 20

Figure 2.2: The luminosity function (LF) (red triangles) for the volume-limited sample is shown, in comparison with the LFs for the much larger sample of Blanton et al. (2003) (blue dots), Bernardi et al. (2013) (green dashed line), and CGS (black stars) (Ho et al. 2011). The LF implies an overdense region for the volume-limited sample, therefore it is renormalized by adding 0.25 on the y-axis. The normalized LF is depicted by pink − circles. Note that the LFs of Blanton et al. (2003) and Bernardi et al. (2013) have been shifted by B r = 0.67 mag, the average color of an Sbc spiral (Fukugita et al. 1995), − which is roughly the median Hubble type of both CGS and the volume-limited sample, −1 −1 and also transformed to H0 = 67.77 km s Mpc . 21

Figure 2.3: Top: Distribution of morphological types in CGS. Bottom: Distribution of morphological types in the volume-limited sample used here. The sample preserves the distribution of morphological types in CGS. 22

Figure 2.4: The T-type histograms for the volume-limited sample (blue) and CGS (red) are shown. The sample preserves the T-type distribution in CGS. The differences between the densities of each T-type are always less than 5%. 23 (2011), as compiled by the NED. Several different band images were used for the mea- surements.

2.2.3 S´ersic Index Measurement

In order to have a reliable S´ersic index measurement for the early-type galaxies, the foreground stars and background galaxies were carefully masked by using SExtractor (Source Extractor; Bertin and Arnouts 1996), and the centers of these galaxies were determined by using the IRAF task IMCNTR. The sky-background flux and its uncer- tainty were estimated from the mean and standard deviation of five median fluxes that were obtained from small boxes near the galaxy-free corners of each image, respectively. The surface brightness profiles were then extracted using the IRAF task ELLIPSE (Tody 1986; Jedrzejewski 1987) with a fixed center and allowing the isophotal position angle and ellipticity to vary. The best S´ersic bulge + exponential disk model for S0 galaxies, and the best S´ersic bulge model for elliptical galaxies were fitted by minimizing χ2 with an iterative procedure. The models were derived three times for each galaxy in order to estimate the S´ersic index error. The uncertainity in the sky-background level was respectively added and subtracted from the surface brightness profile data in the sec- ond and third derivation (see Figure 2.5). This method for estimating the errors on the model parameters was also used by de Jong (1996). When fitting the profiles, seeing effects are particularly relevant when the ratio between the FWHM of the seeing and the effective half-light radii (Re) of the S´ersic model is small (Graham 2001). When

Re/FWHM > 2, the difference between the measured S´ersic index and the actual S´ersic index is typically small, as explained by Graham (2001). For the sample, all the derived bulge values for Re were greater than 1 arcsec, and the ratio Re/FWHM was greater than 2 (see Table 2.1, Column 6). The results of the best-fitting S´ersic bulge model for elliptical galaxies and the best-fitting S´ersic bulge + exponential disk model for S0 24 galaxies are shown in Figures 2.6 and 2.7, respectively.

Figure 2.5: This is the illustration of the methodology for estimating the errors on the model parameters. The galaxy profile (red dots), the galaxy profile + sky standard deviation (black dots), and the galaxy profile sky standard deviation (blue dots) are − shown with their best model fits for each galaxy. The differences between the radial profiles from the observed galaxy and its models are also shown below each panel. Left panel shows the S´ersic bulge models (dashed lines) for the elliptical galaxy (NGC 1439). The red one refers to the S´ersic bulge model with Re = 57.02, n = 4.98, and µe = 22.65, the black one refers to the model with Re = 82.98, n = 5.56, and µe = 23.28, and the blue one refers to the model with Re = 41.34, n = 4.48, and µe = 22.10. Right panel shows the total galaxy models (dashed lines) of S´ersic bulge models (dotted lines) and exponential disk models (dotted-dashed lines) for the (ESO 208- G021). The red dashed line refers to the model with Re = 14.02, n = 2.37, µe = 19.12, Rdisk = 77.77, and µ0 = 22.26, the black one refers to the model with Re = 14.39, n = 2.40, µe = 19.16, Rdisk = 116.23, and µ0 = 22.33, the blue one refers to the model with Re = 12.61, n = 2.20, µe = 18.95, Rdisk = 39.35, and µ0 = 21.44.

The S´ersic index measurements were successfully completed for all 68 galaxies in the sample. Before proceeding, note that Equation 2.1 was constructed in the R-band (Graham and Driver 2007), while the data in this work ranges from the R-band to 4.6µm. The structural parameters of a galaxy may vary with wavelength due to the ra- dial variations in stellar population and/or dust obscuration (Kelvin et al. 2012). This may result in different values for S´ersic index in different wavelengths. However, the local early-type galaxies mostly have fairly small color gradients (e.g., Peletier et al. 25

Figure 2.6: The surface brightness profiles for 30 elliptical galaxies are shown with the best-fit galaxy models. The dashed line is the best S´ersic fit to the bulge. The differences between the radial profiles from the observed galaxy and its model are also shown below each panel. 26

Figure 2.6 (Continued.) 27

Figure 2.7: The surface brightness profiles for 38 S0 galaxies are shown with the best-fit galaxy model (black dashed line). The pink dotted line is the S´ersic fit to the bulge, the blue dotted-dashed line is the exponential disk fit. The differences between the radial profiles from the observed galaxy and its model are also shown below each panel. 28

Figure 2.7 (Continued.) 29

Figure 2.7 (Continued.) 30 1990; Taylor et al. 2005). Using a similar fitting method to that used here (S´ersic bulge model for ellipticals; S´ersic bulge + exponential disk model for disk galaxies), McDonald et al. (2011) found that the S´ersic indices of elliptical and S0 galaxies show no significant variation across optical and NIR wavelengths. In order to quantify how photometric and structural parameters of a galaxy vary with wavelength, recent studies used 2D single S´ersic fits and reported that galaxies with different S´ersic indices and col- ors follow different trends with wavelength (e.g., Kelvin et al. 2012; Vulcani et al. 2014; Kennedy et al. 2015). Their common result is that high-n galaxies remain relatively stable at all wavelengths. These high-n galaxies roughly correspond to the early-type sample in this work. However, it is worth mentioning that the measurement of the S´ersic index in these recent studies differs from that reported here: they used a single S´ersic profile fit for all galaxies and made no attempt to remove objects for which a two-component fit would be more appropriate. Therefore, single S´ersic index wave- length dependence mostly gives information about bulge and disk properties of a galaxy (Kennedy et al. 2016). For example, Vulcani et al. (2014) attribute the lack of varia- tion in S´ersic index with wavelength for red galaxies to the fact that they principally comprise one-component objects (i.e., ellipticals) or two-component galaxies in which the components possess very similar colors, i.e., S0s. Although one can get some insight for the (disk-less) elliptical galaxies, the single S´ersic galaxy model is not suitable for quantifying possible changes with wavelength to S´ersic indices of bulges in S0 galaxies. Therefore, following the work of McDonald et al. (2011), here no correction was applied to the S´ersic index measurements. All measured data for individual early-type galaxies in the sample are listed in Table 2.1. 31 is galactic HST B LCO LCO A Spitzer Spitzer Spitzer Spitzer 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS ), ) Telescope ), converted from the is luminosity distance ⊙ ⊙ 23 23 23 23 17 04 53 12 05 08 27 20 07 07 34 31 23 15 13 35 23 22 02 09 14 53 21 15 51 48 15 29 02 14 32 08 ...... L 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 D +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − HyperLeda M/M 41 41 47 65 12 74 26 38 97 76 69 93 89 42 13 43 57 32 M/M ...... 8 8 7 7 8 6 8 6 7 7 7 8 7 8 7 8 7 6 log( ED), and 92 62 25 25 21 05 94 15 11 17 20 13 19 18 10 06 49 26 22 45 36 28 19 91 26 71 83 43 55 36 58 78 03 17 19 04 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − n 04 05 21 32 29 60 63 38 99 63 53 15 84 91 11 37 35 08 ...... laxy name. (2) Hubble type, from -band magnitude, determined from the formula: 7 2 B 20 4 11 3 14 2 FWHM . (3) Luminosity distance in Mpc, compiled from the / e ute R -band magnitude (taken from les imaging data was taken. B µm µm µm µm Waveband 14 H 42 4 11 3.6 68 3.6 71 H 23 2 62 3.6 12 H 9 1 30 F160W 41 3 52 R 5 1 93 3.6 55 H 10 2 54 H 25 2 62 H 13 7 68 H 3 2 24 H 6 1 13 H 30 4 57 R 14 2 38 H 2 1 07 H 29 4 ...... B is total apparent 20 20 19 21 19 20 19 19 19 20 20 19 19 20 19 19 19 20 M B − − − − − − − − − − − − − − − − − − m FWHM ratio. (7) Index. S´ersic (8) BH mass in log( / L e R D , where B A − )+5 L -band (from Schlafly and Finkbeiner 2011, as compiled by the N D G012 S0/a 18.3 G026 E 12.4 G021 S0* 17.0 B − − − . (5) Waveband. (6) 5 log( pc − B NGC 1340 E 18.2 NGC 1332 E* 18.9 NGC 1326 S0/a 16.9 NGC 1316 S0 19.2 NGC 1302 S0/a 20.0 NGC 1291 S0/a 8.6 NGC 1172 E 24.3 NGC 1201 S0* 20.2 NGC 1052 E 19.6 NGC 936 S0/a 20.7 NGC 720 E 23.9 NGC 636 E 25.4 NGC 596 S0* 20.6 IC 5181 S0 24.8 ESO 311 ESO 221 NGC 584 E 19.5 Galaxy Name Type ESO 208 (1) (2) (3) (4) (5) (6) (7) (8) (9) m = B mean redshift-independent distance from the NED. (4) Absol in units of M http://cgs.obs.carnegiescience.edu/CGS/database_tab Table 2.1: Early-type sample for local BHMF. Columns: (1) Ga extinction in the ´ri indexS´ersic via Equation 2.1. (9) Telescope from which the 32 HST HST LCO LCO LCO Spitzer Spitzer Spitzer Spitzer 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS ) Telescope ⊙ 09 10 04 25 08 11 45 13 30 34 15 06 13 28 07 07 24 28 13 33 39 24 06 44 05 07 10 11 09 72 20 26 20 75 15 17 14 18 08 22 04 02 08 33 16 37 20 10 15 15 17 29 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − M/M 72 56 24 26 82 77 28 09 14 57 92 90 32 65 91 51 73 76 65 85 12 11 65 82 33 32 ...... 7 8 8 7 7 6 8 6 7 8 7 7 7 8 5 8 7 8 8 6 8 6 8 7 8 8 log( 14 13 18 84 21 27 53 12 56 46 11 04 39 63 04 03 23 22 63 07 86 37 11 62 05 07 58 50 04 26 91 80 32 81 21 91 83 76 06 15 10 05 04 46 97 34 35 15 47 40 54 69 ...... 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − n 57 59 56 02 72 68 91 63 89 86 09 98 15 39 62 02 68 62 28 24 23 03 73 82 79 78 ...... 5 1 3 1 9 1 10 2 FWHM / e R µm µm µm µm Waveband Early-type sample. (continued) 17 F160W 32 2 46 H 12 4 39 H 14 3 29 K 12 2 56 H 3 2 11 H 33 3 49 J 6 1 24 4.5 10 H 34 4 82 H 16 2 56 3.6 81 F160W 21 2 65 R 56 4 78 H 7 1 22 H 45 4 43 H 23 5 80 R 52 5 05 H 8 1 49 3.6 57 4.5 54 H 10 3 04 H 5 1 23 H 14 5 82 R 54 2 13 H 26 3 22 H 15 3 ...... B 19 19 19 20 19 21 19 19 21 19 19 19 19 19 21 20 19 20 20 20 20 20 20 19 20 19 M − − − − − − − − − − − − − − − − − − − − − − − − − − L D (1) (2) (3) (4) (5) (6) (7) (8) (9) Galaxy Name Type NGC 1351 E* 20.4 NGC 1374 E 19.4 NGC 1379 E 18.1 NGC 1380 S0 18.0 NGC 1387 S0* 18.0 NGC 1395 E 21.9 NGC 1527 S0* 16.7 NGC 1452 S0/a 22.8 NGC 1399 E 18.9 NGC 1537 E* 18.8 NGC 1533 S0* 18.4 NGC 1400 E* 23.6 NGC 1439 E 24.3 NGC 1543 S0 17.5 NGC 1407 E 23.8 NGC 1549 E 16.4 NGC 1427 E 20.9 NGC 1574 S0* 18.6 NGC 1404 S0* 18.6 NGC 1553 S0 14.6 NGC 2380 S0 22.2 NGC 2217 S0/a 19.5 NGC 2434 E 21.9 NGC 2325 E 22.6 NGC 2640 E* 17.2 NGC 2784 S0 8.5 33 LCO LCO LCO LCO LCO Spitzer Spitzer Spitzer Spitzer Spitzer Spitzer Spitzer Spitzer 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS 2MASS ) Telescope ⊙ 11 03 12 05 07 46 02 45 05 23 15 19 30 12 14 16 18 19 47 22 18 19 26 05 08 09 25 28 46 44 14 14 56 70 15 10 17 11 23 19 16 30 07 30 01 04 35 30 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − M/M 47 35 67 53 82 46 10 67 86 59 99 04 19 34 34 60 65 79 10 61 74 63 74 07 ...... profile. 7 8 7 8 5 8 8 7 7 8 7 8 8 8 7 8 6 8 8 7 7 7 6 7 log( 13 04 38 15 10 53 09 27 02 09 61 58 77 23 20 20 33 29 20 80 38 32 60 09 20 20 91 68 58 39 74 56 44 37 33 64 40 22 34 23 25 39 10 35 01 03 34 23 ...... 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +5 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − n 25 86 50 12 22 23 79 74 54 02 12 44 85 11 77 46 51 83 23 42 60 45 60 85 ...... 9 4 8 2 10 2 16 3 12 2 10 1 20 3 11 2 FWHM / e R µm µm µm µm µm µm µm µm Waveband 4H 42 Early-type sample. (continued) 20 J 58 4 36 K 13 2 34 H 12 4 53 H 20 1 00 H 47 3 62 4.5 13 H 19 3 38 H 7 3 03 3.6 63 H 35 3 74 3.6 46 3.6 15 3.6 83 3.6 28 3.6 93 R 92 4 09 3.6 35 K 16 5 30 R 18 3 35 H 3 2 32 R 11 2 62 R 6 1 81 R 7 1 ...... B 19 20 20 19 20 20 19 21 19 20 20 19 20 19 19 20 20 20 20 20 20 19 20 19 M − − − − − − − − − − − − − − − − − − − − − − − − L D * Hubble type for this galaxy is determined based on its light NGC 4976 E 12.5 NGC 4856 S0/a 21.1 NGC 4024 E* 25.4 NGC 5128 S0 3.7 NGC 4697 E 11.6 NGC 4984 S0* 21.3 NGC 3923 E 20.9 NGC 4684 S0/a 20.5 NGC 4958 S0 20.9 NGC 3585 E 17.6 NGC 4546 S0* 17.3 NGC 4753 S0 19.6 NGC 3955 S0/a 20.6 NGC 4691 S0/a 22.5 NGC 3904 E 24.7 NGC 3136 E 23.9 NGC 3115 S0* 10.1 NGC 7507 E 22.5 (1) (2) (3) (4) (5) (6) (7) (8) (9) NGC 2974 S0* 25.3 NGC 7144 S0* 24.9 NGC 6684 S0 12.4 NGC 7145 S0* 23.1 Galaxy Name Type NGC 2822 S0* 24.7 NGC 7041 S0* 24.9 34 2.2.4 S´ersic Index Distribution

As a result of S´ersic index measurements, three S´ersic index estimates (ni) are obtained for each of 68 E and S0 galaxies. Two independent methods were used in order to find the best-fit probability density function (PDF) to the data. First, a nominal binless histogram, which is identical to the method in Davis et al. (2014), was employed in order to create the S´ersic index distribution. Each data point was modelled twice as a normalized Gaussian, where the mean was the average S´ersic index values < ni > and the standard deviation was the standard deviation of ni ,

σ. The S´ersic index distribution was obtained by a normalized sum of Gaussian values. The same modeling was then repeated, but this time the mean was the aver- age logarithmic value of ni, < log(ni) >, and the standard deviation was the standard deviation of log(ni), σ. From the resulting S´ersic index distributions, the sta- tistical standardized moments of a probability distribution were computed; mean (µ), standard deviation (stdev), skewness, and kurtosis. The two distributions gave al- most the same statistical standardized moments: µ = 3.10(3.10), stdev= 1.38(1.39), skewness= 0.95(0.95), and kurtosis = 4.17(4.18), where the numbers in parentheses refer to the distribution derived from < ni > and σ. The MATLAB code PEAR- SPDF was used to perform the PDF fitting. To explore the uncertainty in the PDF fit, a bootstrapping process was used. The random number generator NORMRND in MATLAB was used for sampling (with replacement) from the original 68 data points, using the mean as < log(ni) > and the standard deviation as σ. The statistical standardized moments for 1000 data sets containing 68 data points each were individ- ually calculated. This gave 1000 new estimates for each of the parameters (µ, stdev, skewness, and kurtosis). The median and the standard deviation of these new estimates then gave the uncertainty on the PDF fitting: µ = 3.12 0.02, stdev= 1.40 0.04, ± ± skewness= 0.92 0.03, kurtosis= 3.87 0.30. ± ± 35

Figure 2.8: The S´ersic index histogram (blue dashed line) and the PDF fits (red solid line from PEARSPDF, pink solid line from ALLFITDIST) to the data are shown. The PDF (red solid line) from PEARSPDF is defined by the statistical standardized moments: µ = 3.10, stdev= 1.38, skewness= 0.95, and kurtosis= 4.17. The PDF (pink solid line) from ALLFITDIST is a gamma distribution function with µ = 3.11, variance= 1.84, shape a = 5.26, and scale b = 0.59. 36 The MATLAB code ALLFITDIST was then used. It was used to fit all valid para- metric probability distributions to the data and returned the fitted distributions based on the Bayesian information criterion. As a result, the gamma distribution function was given as a best PDF fit, with µ = 3.11, variance= 1.84, shape a = 5.26 0.51, ± scale b = 0.59 0.06. The resulting S´ersic distribution and its PDF fits are illustrated ± in Figure 2.8.

2.2.5 Estimating BHMF

The local BHMF is formulated as

∂N ∂N ∂x ∂x φ(log(MBH /M⊙)) = = = φ(x) ∂ log(MBH /M⊙) ∂x ∂ log(MBH /M⊙) ∂ log(MBH /M⊙) (2.2) where N is the number of galaxies, x is pitch angle P for late-type galaxies and S´ersic index n for early-type galaxies. For the early-type galaxies, the S´ersic index measure-

∂N ments for the volume-limited sample give the S´ersic index function φ(n) = ∂n ; and ∂n can be evaluated by taking the derivative of Equation 2.1 as follows: ∂ log(MBH /M⊙)

n d log(M /M⊙) (3.70 0.46) 2(3.10 0.84) log( ) BH = ± ± 3 (2.3) dn n ln(10) − n ln(10)

As a result, the following equation is obtained:

n (3.70 0.46) 2(3.10 0.84) log( 3 ) −1 φ(log(M /M⊙)) = φ(n)[ ± ± ] (2.4) BH n ln(10) − n ln(10)

Using Equation 2.4 and dividing by a local comoving volume of V = 3.37 104 h−3 c × 67.77 Mpc3, the S´ersic index distribution was converted into the BHMF for the early-type galaxies. In order to estimate the error in the BHMF, a Markov Chain Monte Carlo (MCMC) sampling of the BHMF was run. The sampling used 105 realizations of the S´ersic index distribution based on the errors in the previous section. The S´ersic index distributions 37

Figure 2.9: Left: Impact of the uncertainties on the shape of the BHMF is shown, first assuming no errors, then allowing the listed errors to be perturbed individually and then collectively. The uncertainities in the S´ersic index distribution have no impact on the 9 BHMF for M > 10 M⊙. The uncertainties in the M n relation dominate at this BH BH − region, softening the high-mass decrease of the BHMF, and thus increasing the total density of the BHMF for high masses. Right: Best estimate of the early-type BHMF is obtained by merging all the MCMC realizations of the BHMFs after allowing the listed errors to be perturbed collectively. The solid red line represents the 50th percentile and the green shaded region is delimited by the 16th and 84th percentile levels. Note that the BHMF estimates are all normalized by adding 0.25 on the y-axis to be able correct − for the overdensity in the survey volume. were randomly generated from the parameters that define the PDF, assuming that they are normally distributed within the 1σ uncertainties given by the estimated errors. The uncertainties in the M n relation were also allowed to vary as a Gaussian distribution BH − around the fiducial values. The BHMF was first estimated without assuming any errors, then the listed errors were allowed (four parameters in the PDF fit + three parameters in the M n relation) to be perturbed individually and collectively. This is illustrated BH − in Figure 2.9 (left), which shows that the S´ersic index distribution had no impact on the

9 BHMF for MBH > 10 M⊙ since the BH mass is fixed for n> 11.9. The sharp decrease at the high-mass end is the result of the curved nature of the M n relation, that predicts BH − a maximum mass that BHs have formed (Graham et al. 2007). The uncertainties in the 38 M n relation dominate at this region, softening the high-mass decrease of the BH − BHMF, and thus increasing the total density of the BHMF for high masses. The error region in the BHMF was estimated by the 16th and 84th percentile of the 105 MCMC realizations, similar to the method used by Marconi et al. (2004), where the 16th and 84th percentiles indicate the 1σ uncertainties on the logarithm of the local BHMF. In order to deal with the intrinsic scatter in the M P relation, Davis et al. BH − (2014) used the method described in Equation (3) of Marconi et al. (2004). However, this method was not adopted here for the early-type BHMF. Graham et al. (2007) posited that the intrinsic scatter in the M n relation is not Gaussian, and the BH − removal of the two highest-mass BHs converts the M n relation into one with BH − zero intrinsic scatter. In estimating the BHMF derived from the M n relation, BH − Graham et al. (2007) did not apply any correction for the intrinsic scatter, and neither did I. Finally, the best estimate of the early-type BHMF was obtained by merging all of the random realizations of the BHMFs and considering the 16th, 50th, and 84th percentile levels (see the right panel in Figure 2.9). Note that the early-type BHMF was normalized by adding 0.25 on the y-axis, which corrected for the overdensity in − the selected volume. In order to estimate the local BHMF for all galaxy types, following Equation 2.2, the MCMC realizations of the BHMF for the spiral galaxies were also run- this time using the pitch angle distribution derived by Davis et al. (2014). Note that Davis et al. (2014) estimated possible BH masses from the M P relation by using the MCMC BH − sampling and then fitted a PDF model to derive the late-type BHMF. In this work, I used the best-fit PDF model for the pitch angle distribution derived by Davis et al. (2014), and then used Equation 2.2 by adopting the method of Marconi et al. (2004) to estimate the late-type BHMF by considering the 16th, 50th, and 84th percentile levels of the MCMC realizations. Similar to the early-type MCMC sampling, it was 39

Figure 2.10: Best estimate of the BHMF was obtained by merging all the MCMC realizations of the BHMFs from the early- and late-type galaxies. The M n and BH − M P relations were used for the early- and late-type galaxies, respectively. The BH − MCMC sampling was used to account for the uncertainies from both the measurements and the scaling relations. The all-type BHMF (red solid line) was defined by the 50th percentile, while its error region (green shaded region) was delimited by the 16th and 84th percentile levels of the merged MCMC realizations. The blue and pink dotted lines show the 1σ uncertainity region for the late- and early-type BHMF, respectively. This clearly shows that the late-type BHMF dominates at the low-mass end while the early-type BHMF dominates at the high-mass end. Note that the BHMF estimates were all normalized by adding 0.25 on the y-axis. − 40 assumed that the input parameters (µ, stdev, skewness, kurtoisis) of the PDF fit and the uncertainities in the M P relation are Gaussian distributed around the fiducial BH − values. All random realizations of BHMFs were then merged from the early-type and spiral galaxies. Figure 2.10 shows the best estimate of the local BHMF obtained by merging all random realizations and considering the 16th, 50th, and 84th percentile levels. The late-type BHMF and the early-type BHMF are also shown in Figure 2.10 to help visualize how the early- and late-type samples were being spliced. Note that the BHMF estimates were all normalized by adding 0.25 on the y-axis to enable correction − for the overdensity in the survey volume. The plotted data for Figure 2.9 (right) and Figure 2.10 are listed for convenience in Table 2.2.

2.2.6 BH Mass Density

Integrating over the mass functions, the local mass density of BHs was derived, giving

+0.79 5 3 −3 +1.16 5 3 −3 1.74 10 h M⊙ Mpc for early-type and 2.04 10 h M⊙ Mpc for −0.60 × 67.77 −0.75 × 67.77 all-type galaxies. For reference, Graham et al. (2007) and Vika et al. (2009) reported

5 3 −3 5 3 −3 3.99 1.54 10 h M⊙ Mpc and 7.25 1.18 10 h M⊙ Mpc for the BH ± × 67.77 ± × 67.77 mass density for all-type local galaxies, respectively. In terms of the critical density of the universe, Ω = 1.61+0.91 10−6 h was obtained. This implies that 0.007+0.005 BH −0.59 × 67.77 −0.003 3 h67.77 percent of the baryons are contained in BHs at the centers of galaxies in the local universe (see Table 2.3).

2.2.7 Discussion

Figure 2.11 shows the comparison of the early-type BHMF derived here with previ- ously estimated early-type BHMFs (Graham et al. 2007; Marconi et al. 2004; Vika et al. 2009). The early-type BHMF derived here is expected to be consistent with that of Graham et al. (2007) within the uncertainties, because they are both derived from the 41 same M n relation. The data points are in overall good agreement within their BH − 6.5 uncertainties. There is an apparent disagreement below MBH < 10 M⊙, which corre-

8 8.75 sponds to n 1.5 and the region between 10 M⊙ < M < 10 M⊙. Graham et al. ≈ BH (2007) defined early-type galaxies as B > 0.4 and used the GIM2D derived n values T − (Allen et al. 2006) that were obtained from the logical filter for S´ersic + exponential catalog. For galaxies with n< 1.5, this logical filter classifies galaxies as pure disk and therefore fits them with a single component. However, I obtain 1

6.5 higher density for the low-mass end (MBH < 10 M⊙) and lower density for interme-

8 8.75 diate masses (10 M⊙ < MBH < 10 M⊙). Differences in the definition of early-type galaxies and the profile fitting methodology may explain the disagreements between the two BHMFs derived from the same relation. It should also be noted that Graham et al. (2007) used a sample of 1356 early-type galaxies from the Millennium Galaxy Cata- logue (MGC) in the redshift range of 0.013

φ(M) = P W (L)M, where W (L) = φ(L)/N(L) is constructed for black holes derived B from early-type galaxies (defined as T > 0.4). Although the volume of their sample is considerably higher than that used here, and their sample selection and BHMF estima- tion method are different from that used here, overall their BHMF is consistent with my findings. Vika et al. (2009) used the sample identical to that of Graham et al. (2007), except they included the galaxies with M > 18, indicating the data from this region is B − unreliable. They used the linear M L relation reported by Graham (2007) BH − Bulge with dust correction to their sample. Other than using the M L relation to BH − Bulge derive the BHMF, their BHMF estimation method is identical to that of Graham et al. (2007). However, their BHMF does not agree well with that of Graham et al. (2007), 42

Figure 2.11: Comparison of the early-type BHMF (solid red line) with a green shaded 1σ error region; with those of Graham et al. (2007) (pink triangles: GR07); ± Marconi et al. (2004) (black stars: M04); and Vika et al. (2009) (blue open circles: V09). The BHMF data of Vika et al. (2009) below log(MBH /M⊙) = 7.67, which Vika et al. (2009) considers unreliable because it is derived from galaxies with MB > 18, is depicted by the open circles with light blue color. Note that the BHMF is − normalized by adding 0.25 on the y-axis to be able correct for the overdensity in the − −1 −1 sample and all other BHMFs are transformed to H0 = 67.77 km s Mpc . 43 or with that reported here. They discussed the probable reasons for the discrepancy between theirs and that of Graham et al. (2007) (see Section 3.1 in Vika et al. 2009). In addition, Graham and Scott (2013) recently revised the M L relation and BH − Bulge found a log quadratic nature for the M L relation, which is also expected BH − Bulge from the linear nature of the two distinct L n relations for elliptical galaxies and Bulge − bulges, and the curved M n relation. This may explain the discrepancy between BH − the BHMF derived from the linear M L relation and the one derived from the BH − Bulge curved M n relation. BH − Marconi et al. (2004) estimated the local BHMF for early-type galaxies based on the SDSS sample of Bernardi et al. (2003), by using the linear M L and BH − Bulge M σ relations reported by Marconi and Hunt (2003) assuming the same intrinsic BH − ⋆ dispersion. They also derived the local BHMF for early-type galaxies obtained from different galaxy luminosity functions, in different photometric bands. All of their local BHMFs for early-type galaxies are in remarkable agreement with the BHMF reported

8 here within the uncertainities. However, they reported a discrepancy at MBH < 10 M⊙ between the BHMF derived with the Bernardi et al. (2003) luminosity function and the others (see Figure 1(b) in Marconi et al. 2004). They considered this discrepancy to be insignificant because this is the region where authors adopted different functional forms to fit the data to extrapolate luminosity functions of early-type galaxies. The early-type BHMF reported here agrees more with the one derived from the sample of

8 Bernardi et al. (2003) at MBH < 10 M⊙ than the others. Figure 2.12 shows the comparison between the BHMF reported here for all galaxy types with those of Graham et al. (2007), Vika et al. (2009), and Marconi et al. (2004). Overall, the BHMF reported here agrees better with that of Marconi et al. (2004) within the uncertainties. It is clear that there is a disagreement between that reported here and those of Graham et al. (2007) and Vika et al. (2009) at the low-mass end. Late-type 44

Figure 2.12: Comparison of the BHMF (red solid line) for all galaxy types with a green shaded 1σ error region; with those of Graham et al. (2007) (pink triangles: ± GR07); Marconi et al. (2004) (black stars: M04) ; and Vika et al. (2009) (blue open circles: V09). The BHMF data of Vika et al. (2009) below log(MBH /M⊙) = 7.67, which Vika et al. (2009) considers unreliable because it is derived from galaxies with M > 18, is depicted by the open circles with light blue color. Note that the BHMF B − is normalized by adding 0.25 on the y-axis to be able correct for the overdensity in − −1 −1 the sample and all other BHMFs are transformed to H0 = 67.77 km s Mpc . 45

Figure 2.13: Comparison of the BHMF (red solid line) for all galaxy types with a green shaded 1σ error region with more recent works. The blue solid lines show the ± 1σ uncertainity region for the local BHMF from Shankar (2013)(S13), assuming the revised M σ∗ relation from McConnell and Ma (2013) and applying it to all local BH − galaxies. The region enclosed by the blue dashed lines is the same, but this is assuming the BH mass in Sa galaxies is negligible. The black dotted-dashed line shows the local BHMF derived by using the continuity equation models of Shankar et al. (2013) and assuming a characteristic Eddington ratio decreasing with cosmological time. The pink dotted line marks the local BHMF in the Illustris simulated volume (Sijacki et al. 2015). Note that the BHMF is normalized by adding 0.25 on the y-axis to be able correct for − the overdensity in the sample and all other BHMFs are transformed to H0 = 67.77 km s−1 Mpc−1. 46 galaxies have the biggest contribution on the BHMF at the low-mass end (see Figure 2.10), where the S´ersic index is more difficult to measure due to the complex nature of these late-type galaxies as explained earlier in this chapter. It is also worth mentioning that Vika et al. (2009) argued that their BHMF data below log(MBH /M⊙) = 7.67 (light blue circles in Figure 2.12) is not reliable because it is derived from galaxies with MB > 18. The entire sample used here consists of galaxies with M 19.12. Moreover, − B ≤ − Davis et al. (2014) stated a possible bias for the sample of Vika et al. (2009), pointing to the small number of late-type galaxies in their considerably larger sample volume (see Section 7 in Davis et al. 2014). Although the sample used here does not contain very faint galaxies (M > 19.12), my BHMF results in a higher number density for the B − low-mass end when compared to those of Vika et al. (2009) and Graham et al. (2007). In addition, other relations (M n, M L , and M σ ) are not as accurate BH − BH − Bulge BH − ⋆ as the M P relation in this mass regime (Berrier et al. 2013). BH − Finally, Figure 2.13 shows the comparison between the all-type BHMF reported here with more recent BHMF estimates (Shankar 2013; Sijacki et al. 2015). At the high-mass end, it looks as if my BHMF lies between those of Marconi et al. (2004) and

7 Shankar (2013), except for the lower-mass BHs with MBH < 10 M⊙. Shankar (2013) derived the local BHMF based on the assumption that all local galaxies follow the early- type M σ relation reported by McConnell and Ma (2013). As shown in Figure BH − ⋆ 2.10, early-type galaxies dominate at the high-mass end, therefore a BHMF derived from a relation for early-type galaxies is expected to be more reliable at the high-mass end. Observational uncertainities increase for low-mass (late-type) galaxies, because measuring σ⋆ in disk galaxies is not a trivial task and one needs to properly count the contributions from the motions of disk and bar that are coupled with the bulge. In addition, the majority of low-mass galaxies may host pseudobulges (Fisher and Drory 2011), and a number of independent groups claimed that the properties measured for 47

Table 2.2: BHMF values. Columns: (1) BH mass listed as log(M/M⊙) in 0.25 dex intervals. (2) Normalized BHMF data for early-type galaxies in the sample, as presented 3 −3 −1 in Figure 2.9 (left), in units of h67.77 Mpc dex . (3) Normalized BHMF data for all 3 −3 −1 galaxies in the sample, as presented in Figure 2.10, in units of h67.77 Mpc dex .

3 −3 −1 log MBH /M⊙ log ϕ [h67.77 Mpc dex ] (1) Early Type (2) All Galaxies (3) 5.00 4.42+0.27 3.84+0.30 − −0.31 − −0.39 5.25 4.29+0.25 3.66+0.24 − −0.31 − −0.35 5.50 4.18+0.24 3.48+0.22 − −0.28 − −0.31 5.75 4.08+0.22 3.30+0.19 − −0.26 − −0.25 6.00 3.96+0.20 3.14+0.14 − −0.24 − −0.22 6.25 3.84+0.17 3.00+0.11 − −0.22 − −0.18 6.50 3.73+0.15 2.88+0.07 − −0.19 − −0.13 6.75 3.62+0.13 2.80+0.06 − −0.17 − −0.08 7.00 3.51+0.11 2.74+0.06 − −0.14 − −0.07 7.25 3.41+0.08 2.72+0.08 − −0.12 − −0.11 7.50 3.32+0.07 2.75+0.12 − −0.09 − −0.15 7.75 3.25+0.06 2.85+0.16 − −0.06 − −0.17 8.00 3.21+0.07 2.97+0.17 − −0.06 − −0.16 8.25 3.20+0.08 3.09+0.16 − −0.09 − −0.15 8.50 3.27+0.11 3.25+0.15 − −0.16 − −0.18 8.75 3.45+0.17 3.44+0.17 − −0.27 − −0.28 9.00 3.71+0.25 3.70+0.25 − −0.39 − −0.39 9.25 4.02+0.35 3.99+0.34 − −0.46 − −0.50 9.50 4.32+0.39 4.32+0.42 − −0.64 − −0.59 48

Table 2.3: BHMF evaluation. Columns: (1) Number of galaxies. (2) Total mass from 10 the summation of all the BHs in units of 10 M⊙. (3) Density of BHs in units of 5 3 −3 10 h67.77 M⊙ Mpc . (4) Cosmological BH mass density [ΩBH = ρ/ρ0, assuming 2 11 −3 −1 −3 ρ 3H /8πG = 1.274 10 M⊙ Mpc when H = 67.77 km s Mpc ]. (5) 0 ≡ 0 × 0 Fraction of the universal baryonic inventory locked up in BHs [ΩBH /ωb].

N MT otal ρ ΩBH ΩBH /ωb 10 5 3 −3 −6 3 (10 M⊙) (10 h67.77 M⊙ Mpc ) (10 h67.77) (h67.77 %) (1) (2) (3) (4) (5) b +0.47 +1.40 +1.10 +0.005 68 (E/S0) 1.05−0.36 3.10−1.06 2.44−0.83 0.011−0.004 a +0.26 +0.79 +0.62 +0.003 68 (E/S0) 0.59−0.20 1.74−0.60 1.37−0.47 0.006−0.002 *,b +0.22 +0.65 +0.51 +0.003 140 (Spiral) 0.18−0.09 0.55−0.27 0.43−0.21 0.002−0.002 a +0.12 +0.37 +0.29 +0.002 140 (Spiral) 0.10−0.05 0.30−0.15 0.24−0.12 0.001−0.001 b +0.69 +2.05 +1.60 +0.008 208 (All-type) 1.23−0.45 3.65−1.33 2.87−1.11 0.013−0.006 a +0.38 +1.16 +0.91 +0.005 208 (All-type) 0.69−0.25 2.04−0.75 1.61−0.59 0.007−0.003

b Before the normalization is applied. a After the normalization is applied. * Data taken from Davis et al. (2014). 49 galaxies with pseudobulges do not follow the typical scaling relations (e.g., M σ , BH − ⋆ M M , M L ), with BH masses being often significantly smaller BH − ⋆,Bulge BH − Bulge than what is anticipated by these relations (e.g., Greene et al. 2010; Kormendy et al. 2011; Beifiori et al. 2012). Therefore, the BHMF of Shankar (2013) (and most previous works) likely represents an upper limit on the true local BHMF (Shankar 2013). To adress this issue, Shankar (2013) re-estimated the BHMF with the same relation, but this time the authors made the odd assumption that Sa galaxies do not host any BHs. This assumption likely makes this modified BHMF a lower limit on the local BHMF (Sijacki et al. 2015). My BHMF indeed stays between the BHMF of Shankar (2013) and the modified one. In the comparison with the BHMFs derived from accretion models, the continuity equation models of Shankar et al. (2013) predict a local BHMF similar to that of Shankar (2013) when a constant Eddington ratio is assumed (see Figure 2 of Shankar 2013), and they predict a local BHMF very similar to that reported here for the highest mass regime when a Eddington ratio is assumed to be decreasing as a function of cosmological time (see dotted-dashed line in Figure 2.13). Finally, when compared with the Illustris Simulation, which is a large-scale cosmological simulation with the resolution of a (106.5 Mpc)3 volume, my result agrees quite well with their BHMF. At higher masses, the simulation estimate is in a remarkable agreement with my result. Similar to the others, disagreements exist at lower masses, and Sijacki et al. (2015) already argued that the simulation results are least reliable at the low-mass end (see Section 3.3 in Sijacki et al. 2015). In summary, for the intermediate- and high-mass

7 BHs (MBH > 10 M⊙), the agreements between the BHMF reported here and those of previous BHMF estimates are encouraging. At the low-mass end, inconsistencies exist in the previous works that still need to be resolved, but my work is more in line with the expectations based on accretion models (Shankar et al. 2013), favouring steadily decreasing Eddington ratios, and semi-analytic models (e.g., Marulli et al. 2008), which 50 8 suggest a relatively flat distribution for MBH . 10 M⊙. The results shown here for the low-mass end of the BHMF are probably consistent with the claims that the majority of low-mass galaxies contain pseudobulges rather than classical bulges (Fisher and Drory 2011). This, in turn, may explain why the M P produces a tighter relation than BH − the M σ relation for disk galaxies (Berrier et al. 2013), and therefore why my BH − ⋆ BHMF result shows more promise when compared to expectations from semi-analytical models. This also highlights an important need for properly accounting for the affects of pseudobulges in disk galaxies when determining the local BHMF.

2.3 The Search for IMBHs

2.3.1 Methodology

Graham (2012) dramatically revised the well-known, near-linear relationship relating the BH mass and the host galaxy bulge mass, demonstrating that it is approximately quadratic for the low- and intermediate-mass bulges. Graham and Scott (2013) and Scott et al. (2013) built on the new M M relation, and by including many BH − ⋆,Bulge more low-mass spheroids with directly measured BH masses, they showed that the M L relation is better described by two power-laws (see also Graham and Scott BH − Bulge 2015). Using this scaling relation, Graham and Scott (2013) identified 40 lower lumi- nosity spheroids that contain AGN according to the spheroid magnitudes reported by Dong and De Robertis (2006) and appear to have a mass of the central black hole that falls in the IMBH range. Recently, the scaling relation of Graham and Scott (2013) was refined by Savorgnan et al. (2016) using a sample of 66 nearby galaxies with a dynamically measured BH mass. The list of Graham and Scott (2013) motivated an extensive search for IMBH can- didates in LLAGN using a multiple method approach. This method consisted of the 51 F P BH that was re-calibrated by Plotkin et al. (2012) and the latest M galactic − BH − properties scaling relations, i.e., the M M (M : spheroid absolute magnitude BH − sph sph at 3.6µm) relation for spiral galaxies by Savorgnan et al. (2016), the M n relation BH − from Savorgnan (2016), and the M P relation from Davis et al. (2017). The use of BH − these four methods, which are largely independent, to estimate the BH mass provides an invaluable consistency check. In the first part of this campaign, only seven sources were targetted and they were chosen primarily because they had relatively high X-ray luminosities (however, still in the LLAGN regime), and had all been detected in archival X-ray observations. The following section only focusses on my contribution to this collaborative project, which was to estimate the central black hole masses by studying the spiral arm mor- phology. The results of the other methods are briefly summarized in Section 2.3.3.

2.3.2 Spiral Arms Analysis

The optical imaging data used for measuring pitch angles were obtained from the NED. The R-band imaging data were taken from various sources. When given the option, the imaging with the best resolution was used, specifically NGC 628 from the Vatican Advanced Technology Telescope (Vatt), NGC 3507 from the Palomar 48-inch Schmidt, NGC 4470 from the SDSS, NGC 3198 from the Kitt Peak National Observatory (KPNO) 2.1m telescope, NGC 3486 from the Bok Telescope, and NGC 3185 and NGC 4314 from the 1.0m Jacobus Kapteyn Telescope (JKT). The 2D fast Fourier transform (2DFFT), which is a widely used method to obtain pitch angles of spiral arms (Saraiva Schroeder et al. 1994; Gonzalez and Graham 1996; Seigar et al. 2008; Davis et al. 2012; Mart´ınez-Garc´ıa et al. 2014), was used. It allows for decomposition of spiral arms into multiple modes of spiral patterns. The method, described in Saraiva Schroeder et al. (1994), measures both the strength and the pitch 52

Figure 2.14: Left: NGC 628 R-band pitch angle as a function of inner radius for its deprojected image. A stable mean pitch angle is determined for the m = 2 harmonic mode (black solid line) from a minimum inner radius of 30 pixels (12 arcsec) to a maximum inner radius of 200 pixels (80 arcsec), with an outer radius of 440 pixels (176 arcsec). The pitch angle can be seen to decrease dramatically with increasing radius beyond approximately 50% of the outermost radius and becoming chaotic near the edge; this is due to a shrinking annulus of measurement reducing the viability of pitch angle measurement in the outer regions of the galaxy. Right: An overlay image of a 14.98 degrees pitch angle spiral on top of NGC 628. This figure is reproduced from Koliopanos et al. (2017) 53 angle of the various modes between a given inner radius and a given outer radius on a deprojected image. The extension to this method in Davis et al. (2012) eliminates the user defined inner radius and allows the user to quantitatively investigate how logarithmic the spiral arms are. This provides a systematic way of excluding barred regions from the pitch angle measurement annulus. I therefore opt for the method of Davis et al. (2012), which outputs the pitch angle as a function of inner radius. The images were first deprojected to face-on by assuming that a disk galaxy, when face-on, has outer isophotes that are circular. The galaxy axis ratios were determined using SExtractor (Bertin and Arnouts 1996). A harmonic mode was chosen in which pitch angle stays approximately constant over the largest range of radii, a region refered to as the “stable region”. The pitch angle was then measured by averaging over this stable region, with the error being determined by considering the size of the stable radius segment relative to the galaxy radius, as well as the degree to which the stable segment is logarithmic. Errors in the determination of the galaxy center or galaxy axial ratio in the deprojection process are not critical to the measurement of the pitch angle (see Davis et al. 2012, for more detail). An example measurement of the pitch angle for NGC 628 is shown in Figure 2.14. The source is measured to have a position angle of -40.5 degrees and elongation of 1.103 using SExtraxtor. The image is deprojected after Gaussian star subtraction is performed to reduce the noise from the foreground stars. Although star subtraction is a helpful step, the presence of foreground stars does not affect the value of the pitch angle in general (Davis et al. 2012). The plot shows the output of the 2DFFT code, which is the pitch angle as a function of inner radius for the deprojected image. A stable mean pitch angle is determined for the m = 2 harmonic mode from a minimum inner radius of 30 pixels (12 arcsec) to a maximum inner radius of 200 pixels (80 arcsec), with an outer radius of 440 pixels (176 arcsec). Equation (7) in Davis et al. (2012) describes the final error in pitch angle measurement: 54 2 2 1/2 Eφ = ((βσ/λ) + (ǫm) ) , where Eφ is the total pitch angle error, ǫm is the quantized error for the dominant harmonic mode, σ is the standard deviation about the mean pitch angle, β is the distance in pixels from the inner-most stable spiral structure to 90% of the selected outer radius of the galaxy (0.9 r ), and λ is the length of the stable ∗ max range of radii over which the pitch angle is averaged. For NGC 628, the equation yields

Eφ = 2.23 degrees with λ = 170 pixels, β = 366 pixels, σ = 0.95 degrees, and ǫ2 = 0.88 degrees. The final determination of pitch angle is therefore 15.07 2.23 degrees. The ± measurements are also tested with the Spirality code of Shields et al. (2015), which uses template fitting rather than Fourier decomposition to measure pitch angle. The results from these two independent methods agree well within the uncertainties. Finally, an overlay image is generated to double check the measurement by visualizing the pitch angle on top of the galaxy and confirming the result by eye (see Fig 2.14). The same procedure is used for all seven galaxies in the sample. The pitch angle estimations are presented in column 5 of Table 2.4.

2.3.3 Results

The latest M P relation is derived by Davis et al. (2017) using a sample of 43 di- BH − rectly measured BH masses in spiral galaxies. The linear-fit is: log(M /M⊙) = (9.01 BH ± +0.04 0.08)(0.130 0.006) P with intrinsic scatter ǫ = 0.22 dex in log(M /M⊙) and ± | | int 0.05 BH a total absolute scatter of 0.44 dex. The quality of the fit can be described with a correlation coefficient of 0.85 and a p-value of 7.73 1013. − × 55

Figure 2.15: Mass estimates of the central BH of seven LLAGN galaxies (colored points), using the multiple method approach versus spheroid stellar mass (M∗,sph). The green stars come from the S´ersic index, the blue squares from pitch angle and the red circles from the F P BH. The results have been plotted on top of the scaling relations from − Savorgnan et al. (2016) (solid lines: Y on X-axis regression, dashed lines: symmetric regression). The data that the authors used to obtain these relations are shown as triangles with error bars (black/gray for ellipticals/spirals). The empty diamonds are two other IMBH candidates (LEDA87300: e.g., Baldassare et al. 2015; Graham et al. 2016 and Pox52: e.g., Barth et al. 2004; Thornton et al. 2008; Ciambur 2016). This figure is reproduced from Koliopanos et al. (2017). 56 sph M − sph 8 7 8 6 7 8 M ...... BH ) 0 0 0 0 0 0 M ± ± ± ± ± ± BH 0 6 6 0 4 1 ...... M ( estimation in units log BH relation by Davis et al. using the n M ) ⊙ 5 6 4 6 3 6 3 7 5 6 3 6 P ...... 0 0 0 0 0 0 M − BH ± ± ± ± ± ± M 6 7 3 7 8 3 ( BH ...... M log BH − 1 6 1 7 1 7 1 7 . . . . 5 6 6 6 9 ...... FP 2 2 2 2 using the . . . ) 7 7 7 ± ± ± ± ⊙ 9 5 6 5 BH < < < . . . . M estimation in units of M ( by Plotkin et al. (2012). (6) BH log M BH anos et al. (2017). Columns: (1) Galaxy Name. (2) Hubble P ) 8 6 4 5 6 9 5 9 3 7 5 − ...... 0 0 0 0 0 0 0 BH ± ± ± ± ± ± ± F P M 5 0 9 3 3 1 5 ( ...... estimation in units of 23 7 33 6 26 7 . . . using the 10 5 69 7 74 5 70 7 . . . . 2 3 1 6 0 6 3 BH ⊙ ± ± ± M ± ± ± ± M P log 07 54 51 . . . 17 98 59 62 . . . . 15 16 14 − − − relation by Savorgnan (2016). (7) n − BH estimation in units of M BH M using the (1) (2) (3) (4) (5) (6) (7) Galaxy Name Type NGC 628 SA(s)c NGC 3185 SBc NGC 3198 SB(rs)c 27 NGC 3486 SAB(r)c 12 NGC 3507 SB(s)b NGC 4314 SB(rs)a 28 NGC 4470 SAc 11 ⊙ M relation by Savorgnan et al. (2016). of (2017). (5) Type. (3) Spiral arm pitch angle. (4) Table 2.4: Multiple method approach that reported in Koliop 57 For each galaxy, the mass estimations from the M P are listed in Table 2.4, BH − together with the results from the F P BH, M n, and M M . While − BH − BH − sph all the galaxies in the sample harbor lower-mass-range BHs (log(M /M⊙) 6.5 on BH ∼ 5 average), their mass is significantly higher than for IMBHs ( 10 M⊙). This result is ≤ supported by all four methods used and applies to all sources. All mass estimations from all methods are consistent within the 1σ error bars. This is particularly important since all scaling relations and the F P BH have been calibrated primarily in the high-mass − regime, for which direct measurements were available. The combination of these four independent methods provides a powerful tool for predicting BH masses in the regime where their spheres-of-(gravitational) influence cannot be spatially resolved, mitigating against outliers from any one relation. Nevertheless, despite these initial encouraging results, it is worth emphasizing that this is a pilot study that involved only seven sources. The mass estimates for the central BHs of seven LLAGN galaxies (colored points) are plotted in Figure 2.15, against spheroid (bulge) stellar mass, M⋆,sph. The results have been plotted on top of the scaling relations from Savorgnan et al. (2016) (solid lines: Y on X-axis regression, dashed lines: symmetric regression). The data that the authors used to obtain these relations are shown as triangles with error bars (black/gray for ellipticals/spirals). The empty diamonds are two other IMBH candidates (LEDA87300: e.g., Baldassare et al. 2015; Graham et al. 2016 and Pox52: e.g., Barth et al. 2004; Thornton et al. 2008; Ciambur 2016). The green stars come from the S´ersic index, the blue squares from pitch angle and the red circles from the F P BH. −

2.4 Conclusions

This chapter presents two important applications of BH scaling relations: a determina- tion of the BHMF in the local universe (Section 2.2) and a search for IMBHs (Section 2.3). 58 The observational simplicity of the approach presented in Section 2.2 and the use of the statistically tightest correlations with BH mass the S´ersic index for E/S0 galaxies − and pitch angle for spiral galaxies make it straightforward to estimate a local BHMF − through imaging data only within a limiting luminosity (redshift-independent) distance

DL = 25.4 Mpc (z = 0.00572) and a limiting absolute B-band magnitude of MB = 19.12. At the low-mass end, inconsistencies exist in the previous work that still need − to be resolved, but this work is more in line with the expectations based on accretion models (Shankar et al. 2013), favoring steadily decreasing Eddington ratios and semi- analytic models (e.g., Marulli et al. 2008) that suggest a relatively flat distribution for

8 MBH . 10 M⊙. The BHMF derived in this work is of particular interest because it provides reliable data, especially for the low-mass end of the local BHMF. Section 2.3 presents the first results of an ongoing project in search of IMBH can- didates in LLAGN galaxies. The F P BH (re-calibrated by Plotkin et al. 2012), the − recently revised M spheroid relations by Savorgnan (2016) and Savorgnan et al. BH − (2016), and the M P relation by Davis et al. (2017) are used to re-estimate the BH − masses of seven LLAGN, which previous calculations placed in the intermediate mass regime. Although the central BHs in all seven galaxies have relatively low masses, the multiple method approach shows that they are not IMBHs. Furthermore, this is the first time that the consistency of these four methods have been investigated in the low-mass regime. The consistency between the estimations builds confidence in the legitimacy of this approach and the techniques it involves. Encouraged by the favor- able results of this study, the methodology presented here will be used to re-estimate the mass of stronger LLAGN-IMBH candidates in the Graham and Scott (2013) list. It is important to note that the sources in this work were on the high-mass end of the Graham and Scott (2013) list. There are thirteen sources in the Graham and Scott 59 4 (2013) sample that have estimated 13 masses of less than 10 M⊙ and, when us- ∼ ing the Dong and De Robertis (2006) bulge magnitudes and taking into account the M L relation by Savorgnan et al. (2016), are still expected to lie within the BH − Bulge IMBH range. For the following steps of this campaign, the multiple-method approach will be employed to re-calculate the mass of their central black holes. Chapter 3

The Super-massive Black Hole Galaxy Connection in − Illustris

A slightly modified version of this chapter has been submitted to Monthly Notices of the Royal Astronomical Society with the following bibliographic reference: B. Mutlu-Pakdil, M. S. Seigar, I. B. Hewitt, P. Treuthardt, J. C. Berrier and L. E. Koval, submitted to MNRAS.

3.1 Overview

A common interpretation for super-massive black hole (BH) scaling relations is that BHs regulate their own growth and that of their host galaxies through AGN-feedback (e.g., Silk and Rees 1998; Springel et al. 2005; Bower et al. 2006; Croton et al. 2006; Di Matteo et al. 2008; Ciotti et al. 2009; Fanidakis et al. 2011). Therefore, these scal- ing relations have important implications not only for BHs and galaxies but also for

60 61 understanding the effects of AGN feedback. Recent observational studies have significantly revised these scaling relations, and reported that galaxies with different properties (e.g. pseudobulges versus real bulges, barred versus non-barred, or cored versus power-law ellipticals) may correlate dif- ferently with their central BH masses (e.g., Graham 2008; McConnell and Ma 2013; Kormendy and Ho 2013; Gebhardt et al. 2011). Therefore, it is of fundamental impor- tance to investigate the BH galaxy connection from a theoretical point of view and also − investigate if BH galaxy co-evolution occurs for all galaxy types or if different galaxy − types hold weaker or stronger physical links with their BHs. Although several cos- mological simulations have investigated the BH galaxy connection and confirmed the − establishment of the scaling relations (e.g., Sijacki et al. 2007; Di Matteo et al. 2008; Booth and Schaye 2009; Dubois et al. 2012; Hirschmann et al. 2014; Khandai et al. 2015; Schaye et al. 2015), the co-evolution of the BHs and their galaxies in different galaxy types has not been studied in detail. This is due to the difficulty in simulating representative galaxy samples with significant spatial resolution. Recently, the Illustris simulation project (e.g., Vogelsberger et al. 2014; Genel et al. 2014) has successfully simulated representative galaxy samples covering the observed range of morphologies and provided good spatial resolution to resolve the basic structural properties of the host galaxies. This has made it possible to study the galaxy − connection in detail. Illustris is a large-scale and high resolution hydrodynamic simulation, with a broad range of astrophysical processes: gas cooling with radiative self-shielding corrections, energetic feedback from growing BHs and exploding supernovae, stellar evolution and associated chemical enrichment and stellar mass loss, and radiation proximity effects for AGN (see Vogelsberger et al. 2014 for details). Sijacki et al. (2015) presented results on the BH scaling relations relative to the bulge (M⋆,Bulge and σ⋆) from the Illustris 62 simulation. They concluded that BHs and galaxies co-evolve at the massive end, but there is no tight relation with their central BH masses for low-mass, blue and star- forming galaxies. Here, the aim is to study BH scaling relations beyond the bulge, i.e., spiral arm pitch angle, total stellar mass (M⋆,total), dark matter mass (MDM ), and total halo mass (Mhalo) to complement the study of Sijacki et al. (2015). The Illustris predictions are compared with other theoretical results and observational constraints to further understanding of the BH galaxy connection. To do so, Section 3.4.2 focuses − on the Illustris spiral galaxies, studies the face-on spiral arm morphology in multi- wavebands (B, R and K) and explores its connection with BH and DM for spiral galaxies in Illustris. Section 3.3.3 focuses on the redshift z = 0 galaxies with 10.0 < log(M⋆,2R/M⊙) < 13.0, where M⋆,2R is the stellar mass within the 2x stellar half-mass radius, and studies the Illustris predictions relative to the M⋆,total, MDM and Mhalo of the host galaxy. Note that Snyder et al. (2015) presented the Illustris predictions of the M M and M M relationships in the context of their implications BH − ⋆,total BH − DM for galaxy morphology. This chapter, however, aims to first clearly understand these different relations (e.g. which one may be the strongest correlation), especially for low- mass, blue and star-forming galaxies. Section 3.4.4 finally presents the results and draws conclusions.

3.2 The Illustris Spiral Arm Morphology and Pitch Angle Scaling Relations

3.2.1 Spiral Arm Morphology in Multi-wavebands

The Illustris Project consists of hydrodynamical simulations of galaxy formation in a box of 75 Mpc h−1 (= 106.5 Mpc) on the side (Vogelsberger et al. 2014; Genel et al. 2014). This volume was simulated with and without baryons at several resolution levels, 63 here I present results from the simulation with the highest resolution (with NDM =

3 3 1820 DM particles and Nbaryon = 1820 baryon resolution elements) which resolves

6 baryonic matter with mass 1.26 10 M⊙. For all calculations, the cosmology used for × the Illustris simulation was adopted, that is cosmological parameters consistent with the latest Wilkinson Microwave Anisotropy Probe (WMAP)-9 measurements (ΩM =

−1 −1 0.2726, Ωλ = 0.7274, Ωb = 0.0456, σ8 = 0.809, and H0 = 70.4 km s Mpc with h = 0.704). Parameters in Illustris were tuned to match the redshift z = 0 stellar mass and halo occupation functions, and evolving cosmic rate density (Vogelsberger et al. 2014), and Torrey et al. (2015) showed that this parameterization produces galaxy populations that evolve consistently with observations. This section focuses on the z = 0 spiral galaxies in Illustris to investigate the spiral arm morphology in multiple wavebands and explore its connection with BH and DM.

Snyder et al. (2015) defined a non-parametric bulge-strength parameter, F (G, M20), to serve as a rough automated assessment of morphological types for the z = 0 galaxies in Illustris. In this parameter, G, is the relative distribution of the galaxy pixel flux values

(the Gini coefficient) and M20 is the second-order moment of the brightest 20% of the galaxy’s flux (Conselice et al. 2003; Lotz et al. 2004). They found that late-type galaxies reside in the (M < 2, G < 0.6) region at 10.0 < log(M /M⊙) < 11.0. Following 20 − ⋆,2R their definition for late-type galaxies, 5131 galaxies were first selected from their g band − morphology catalog, based on their location in the G M plane. Then, the face-on − 20 synthetic images in the Johnson-B (444.9 nm) filter were used to identify 95 galaxies that display convincing evidence of definable spiral structure from image inspection. The justification for using only 95 galaxies with spiral structure out of the full sample of possible late-type galaxies is twofold. First, the methods of measuring spiral arm pitch angle are currently very much dependent on the visual inspection conducted by the user (Davis et al. 2012). Also, the full sample includes a large number of galaxies 64

Figure 3.1: Example images in the B, R, and K filters of the z = 0 Illustris galaxies are shown from left to right, respectively. Top: ID = 450471, Middle: ID = 41098, Bottom: 10 ID = 339972. They are arranged by increasing stellar mass from top (M⋆,2R 10 M⊙) 11 ∼ to bottom (M 10 M⊙). ⋆,2R ∼ 65 with a “ring-like” morphology or an irregular/disturbed shape (Vogelsberger et al. 2014; Snyder et al. 2015). Therefore, I only attempt pitch angle measurements on 95 galaxies with definable galaxy structure. After identifying the spiral sample, the spiral arm morphology was studied by using the face-on synthetic images in the Johnson-B (444.9 nm), Cousins-R (659.9 nm) and Johnson-K (2.2 µm) filters, which are available at the Illustris webpage (http://www.illustris-project.org/galaxy_obs/) (Torrey et al. 2015). For brevity, only the data reported for the face-on viewing direction, i.e., camera index = 3, was used since the algorithm requires image deprojection as this produces the most accurate results (Davis et al. 2012). Figure 3.1 shows example images in the B, R, and K filters (from left to right) of the z = 0 Illustris galaxies, arranged by

10 11 increasing stellar mass from top (M 10 M⊙) to bottom (M 10 M⊙). The ⋆,2R ∼ ⋆,2R ∼ remainder of this section analyzes measurements of synthetic images that are most like those in this figure. For the measurements, the two-dimensional (2D) fast Fourier transform (FFT) soft- ware called P2DFFT1 is employed. This software is an updated version of 2DFFT (Davis et al. 2012) that implements parallel processing, allowing it to run faster than the serial 2DFFT, and has simplified the user input. See Saraiva Schroeder et al. (1994); Puerari and Dottori (1992); Puerari et al. (2000) for more details on the 2DFFT algo- rithm. P2DFFT has been written to allow direct input of FITS images and optionally output inverse Fourier transform FITS images. Also included in the package is the abil- ity to generate idealized logarithmic spiral test images of a specified size that have 1 to 6 arms with pitch angles of -90◦ to 90◦. P2DFFT further includes a Python code that outputs Fourier amplitude versus inner radius and pitch angle versus inner radius for each Fourier component (0 m 6). This Python routine also calculates the Fourier ≤ ≤ amplitude weighted mean pitch angle across 1 m 6 versus inner radius. ≤ ≤ 1 P2DFFT is available for download at https://treuthardt.github.io/P2DFFT/ 66

Figure 3.2: Left: Example B-band pitch angle profile of a simulated spiral galaxy (ID = 41098) from the sample is shown as a function of inner radius. A stable mean pitch angle is determined for the m = 2 harmonic mode (solid black line): P = 15.61 3.06 ± degrees. Right: An overlay image of a 15.61 degrees pitch angle spiral is shown on top of this simulated galaxy.

Operationally, P2DFFT decomposes galaxy images into logarithmic spirals and de- termines the pitch angle that maximizes the Fourier amplitudes for each harmonic mode (m). This code provides a systematic way of excluding barred nuclei from the pitch an- gle measurement annulus by allowing inner radius to vary, and outputs the pitch angle as a function of inner radius. By using the face-on synthetic images, the deprojection process, and any assumption related with it, is avoided. Then, a harmonic mode was chosen in which pitch angle stays approximately constant over the largest range of radii, and pitch angle was measured by averaging over this stable region. The error was de- termined by considering the size of the stable radius segment relative to the galaxy radius as well as the degree to which the stable segment is logarithmic (see Davis et al. 2012 for more detail). An example of the pitch angle measurements is shown in Figure 3.2: a stable mean pitch angle was determined for the m = 2 harmonic mode from a minimum inner radius of 50 pixels to a maximum inner radius of 80 pixels, with an 67 outer radius of 110 pixels. The same procedure was used for all 95 galaxies in the B, R and K filters. Each galaxy was also classified as barred/non-barred according to the presence/absence of a clear sign of a bar structure. This classification was based on visual inspection of the K-band images since near-infrared (NIR) bands show bars more frequently (Eskridge et al. 2000). A majority of the sample consists of non-barred morphologies: 21 barred and 74 non-barred. The pitch angle measurements along with the data related with this chapter are presented in Table 3.1. It is important to investigate the possibility of different pitch angles arising in differ- ent wavebands of light. While optical B-band images tend to trace the bright massive star-forming regions of a galaxy, NIR images tend to trace the old stellar populations (and thus the spiral density wave) in galaxies (Seigar and James 1998; Eskridge et al. 2002). Moreover, a flocculent spiral structure in the B-band may appear as a grand- design in the NIR imaging (Thornley 1996). Seigar et al. (2006) demonstrated the existence of a 1:1 relation between the B- and NIR-band (either Ks or H) pitch an- gles for a sample of 66 galaxies from a combination of the Carnegie-Irvine Galaxy Survey (Ho et al. 2011) and the Ohio State University Bright Spiral Galaxy Survey (Eskridge et al. 2002). Later, Davis et al. (2012) confirmed this relation (between B and I) by remeasuring a subset of 47 galaxies of the galaxies appearing in Seigar et al. (2006). Note that Seigar et al. (2006) used an earlier version of the 2DFFT method and Davis et al. (2012) used a method similar that used here. Recently, Pour-Imani et al. (2016) found a modest difference between the bands at the extreme ends of this range, from B to 3.6µm, but concluded that their results are consistent with the results of Seigar et al. (2006) and Davis et al. (2012) because of the small difference they recov- ered. Here I carefully investigated the simulated spiral arms when viewed in different wavebands and checked if Illustris establishes a 1:1 relation that is expected from the observations. Figure 3.3 presents the pitch angle measurements in different wavebands, 68

Figure 3.3: The pitch angle measurements are shown in different wavebands (Left: R versus B, Middle: K versus R, Right: K versus B). Galaxies with bars are shown by green stars. Galaxies with no clear sign of a bar are shown by black circles. The total RMS scatter is plotted in dotted red lines above and below the 1:1 relation line (dashed red line): 4.34◦ in R versus B, 5.24◦ in K versus R and 5.60◦ in K versus B. 69 showing that pitch angles are similar whether measured in the optical or NIR regimes. Although there are small-scale differences between spiral arms in different wavebands of the optical-NIR spectrum, the absence of a systematic behavior in pitch angle val- ues (such as being above or below relative to the 1:1 relation line) is consistent with the observational findings of Seigar et al. (2006) and Davis et al. (2012). The overall structure of the simulated spiral arms is consistent across the optical-NIR spectrum. Furthermore, neither barred (green stars) nor non-barred galaxies (black circles) show a significant difference in pitch angle in different wavebands. This implies that pitch angle derived is not biased by a presence of a bar. In addition, the most common K-band pitch angle in the sample is 21◦, which is very close to the most probable pitch angle of 18.52◦ that Davis et al. (2014) derived from a statistically complete collection of the brightest spiral galaxies.

3.2.2 Pitch Angle Scaling Relations

Observational evidence shows a tight correlation between pitch angle and the central BH mass in disk galaxies (Seigar et al. 2006; Berrier et al. 2013). This relation is expected from our understanding of galactic physics. It has been shown empirically that there is a link between pitch angle and the central mass concentration of galaxies (Seigar et al. 2005, 2006, 2014). Also, the BH bulge connection is widely established as a result of − observed correlations of the BH mass with σ⋆, M⋆,Bulge and LBulge. Since both the BH mass and pitch angle are intimately related to the central mass concentration, it is no surprise that they correlate with each other quite strongly. For this relation, Berrier et al. (2013) reported a scatter less than 0.48 dex, which is lower than the in- trinsic scatter ( 0.56 dex) of the M σ relation, using only late-type galaxies ≈ BH − ⋆ (G¨ultekin et al. 2009). While pitch angle depends inversely on central bulge mass in bulge-dominated galaxies, it correlates to relative concentration of mass toward the 70 galaxy’s center in disk-dominated galaxies, especially the extreme case of bulgeless galaxies (Berrier et al. 2013). Unfortunately, the relation between the BH mass and galaxy characteristics in disk-dominated galaxies is not clear empirically since relatively few BH masses have been directly measured in such galaxies.

MBH was adopted directly from the simulation outputs as the BH mass contained within the stellar half-mass radius. For a given simulated galaxy, MDM was defined as the total mass of dark matter and Mhalo as the total mass of all particles (all types) bound to the subhalo. The K-band pitch angle values were used in the scaling relations. Figure 3.4 shows the Illustris prediction (solid blue line) for the BH mass relative to pitch angle, and compared with the observed M P relation (solid red line) by Berrier et al. BH − (2013). The agreement between the Illustris result and the observations is very good: the slope and normalization of the observed log(M /M⊙) P relation are 0.062 BH − − ± 0.009 and 8.21 0.16, respectively, whereas the simulation predicts 0.055 0.001 and ± − ± 8.40 0.01. The Illustris best fit was obtained using the robust linear least-squares ± fitting method (i.e., bisquare weights) in MATLAB Curve Fitting Toolbox (Goodness of fit: R2 = 0.36, RMSE= 0.44). The same fitting method is used throughout the chapter. Figure 3.5 displays the Illustris predictions for DM mass and halo mass relative to pitch angle. The simulation predicts a very tight correlation between these parameters: the slope and normalization are 0.029 0.001 and 12.46 0.01 (R2 = 0.22, RMSE − ± ± = 0.30) for the log(M /M⊙) P relation, respectively, whereas 0.027 0.001 and DM − − ± 2 12.47 0.01 (R = 0.21, RMSE = 0.29) for the log(M /M⊙) P relation. It is ± halo − especially interesting that Illustris predicts a tighter correlation between pitch angle and DM mass (and halo mass) when compared to the relation between pitch angle and the BH mass. This suggests that pitch angle can be used as a proxy for the DM mass (and halo mass) of disk galaxies. 71

Figure 3.4: The Illustris prediction for the BH mass relative to pitch angle is shown by the solid blue line, which represents the best fit to the data. Galaxies with bars are shown by green stars. Galaxies with no clear sign of a bar are shown by black circles. The total RMS scatter of 0.44 dex in the log(M /M⊙) direction is plotted ± BH in dotted blue lines above and below the best-fit line. For comparison, the observed M P relation by Berrier et al. (2013) is also plotted, represented by the solid red line BH − and bounded by its 0.38 dex RMS scatter in the log(M /M⊙) direction, above and ± BH below with dotted red lines. Overall, the Illustris simulation reproduces the observed relation very well for disk galaxies. 72

Figure 3.5: The Illustris predictions for the DM mass and halo mass relative to pitch angle are shown by the solid blue lines, which represent the best fits to the data: left is the M P relation and right is the M P relation. Galaxies with bars are DM − halo − shown by green stars. Galaxies with no clear sign of a bar are shown by black circles. The total RMS scatters are plotted as dotted blue lines above and below the best-fit lines: 0.30 dex in the log(M /M⊙) direction and 0.29 dex in the log(M /M⊙) ± DM ± halo direction. Note that the Illustris predictions for these correlations are tighter than the one between pitch angle and the BH mass.

Based on limited data, Seigar et al. (2014) uncovered a weak trend between pitch angle (used as a proxy for the BH mass) and DM concentration such that galaxies with tightly wound arms have particularly high concentrations. Furthermore, cosmological simulations often reveal an anti-correlation between DM mass and concentration (e.g., Bullock et al. 2001; van den Bosch et al. 2014). Here, it is shown that the Illustris sim- ulation results in an anti-correlation between DM mass and pitch angle (see Figure 3.5). In conjunction with the DM mass versus concentration anti-correlation (Bullock et al. 2001; van den Bosch et al. 2014) and a DM mass versus pitch angle correlation (Figure 3.5), one expects to derive a positive correlation between pitch angle and DM concen- tration. This is opposite to the weak trend presented in Seigar et al. (2014). However, given the large scatter in the DM mass versus concentration relations revealed by the simulations, and the limitations of the observational data discussed by Seigar et al. (2014) no strong conclusions can be made. Indeed, Seigar et al. (2014) commented that 73 more observational data is needed to really test the predictions made in theoretical studies, such that it would be possible to truly determine, observationally, how DM mass and pitch angle are related or, indeed, if they are related. Moreover, it would be worth testing this link with the future Illustris dark matter halo catalogs (Chua et al. 2016) in a future study. Many works (e.g., Graham 2008; Hu 2008; G¨ultekin et al. 2009; Graham and Scott 2013) have identified an offset in the BH bulge relations between galaxies with barred − and non-barred morphologies. For the pitch angle scaling relations, no offset is found between the barred (green stars) and non-barred (black circles) galaxies in Illustris.

3.3 Black Hole Scaling Relations Beyond the Bulge

This section focuses on the z = 0 Illustris galaxies with 10.0 < log(M⋆,2R/M⊙) < 13.0 to study the BH mass scaling relations beyond the bulge, i.e., the M M , the BH − ⋆,total M M and M M relations. M is adopted directly from the simulation BH − DM BH − halo ⋆,total outputs as the total mass of the stellar particles bound to the subhalo. Note that Snyder et al. (2015) also presented these relations as a function of galaxy morphology. Here, I investigated these predictions by comparing them with other theoretical results and observational constraints at z = 0. The aim was to clearly understand these different relations, e.g., which one may be the strongest correlation, whether different galaxy types exhibit weaker/stronger physical links with their BHs. Moreover, a better understanding of the relations beyond the bulge could provide a benchmark for high- redshift studies which cannot avail themselves of bulge masses or dynamical BH masses (Reines and Volonteri 2015). 74 3.3.1 The M M Relation BH − ⋆,total Figure 3.6 displays the Illustris prediction of the M M relation with the BH − ⋆,total solid yellow line, which presents the best fit to the simulation data: log(MBH /M⊙) =

2 (1.51 0.02) log(M /M⊙) (8.58 0.23), with R = 0.74, RMSE = 0.37. For compar- ± ⋆,total − ± ison, I have also plotted the observed M M relations by Reines and Volonteri BH − ⋆,total (2015), which used a sample of 262 local broad-line AGN and 79 galaxies with dy- namical BH masses. Their total stellar mass measurements rely on mass-to-light ra- tios. The authors defined a relation for AGN host galaxies (solid blue line) that has a similar slope to early-type galaxies with quiescent BHs (dashed blue line), but a normalization that is more than an order of magnitude lower. They also suggested that a potential origin for these different relations could be differences in host galaxy properties (e.g., the Hubble types), claiming that spirals/disks tend to overlap with AGN host galaxies. Although the Illustris galaxies lie within the observational rela- tions of Reines and Volonteri (2015), it seems that Illustris under-estimates the BH masses of early-type galaxies for a given M⋆,total when compared to the observational relation (dashed blue line). Moreover, Illustris appears to favor a single linear-relation for all galaxy types, disagreeing with the findings of Reines and Volonteri (2015). To explore the relation based on bar morphology, the barred and non-barred spiral galaxies are depicted by different signs (green dots and red crosses, respectively). The dashed yellow line is the best fit to the spiral sample, which is consistent with the best fit that is obtained from all galaxy types. Note that the apparent scarcity of the galax-

10.5 ies with M 10 M⊙ in the spiral sample is due to the high abundance of ⋆,total ∼ peculiar galaxies, i.e., ringed galaxies, in this mass range (Snyder et al. 2015). A com- parison is also made with the prediction of the high-resolution cosmological simulation MassiveBlackII (DeGraf et al. 2015), showing remarkable agreement with the Illustris prediction with a slightly shallow slope (see the magenta line). Their relation, which 75 is derived for all galaxy types, is in remarkable agreement with the best-fit for the spi- ral sample. This supports the idea that BHs in spiral galaxies correlate with M⋆,total in a similar way to those in early-type galaxies do. DeGraf et al. (2015) showed that the low-end of their M σ relation tends to lie above the observations, which is BH − ⋆ consistent with Sijacki et al. (2015), who found similar behavior at the low-end. How- ever, the authors showed that their M M relation shows good agreement BH − ⋆,total with the observations at the high-end as well as at the low-end of the relation. Based on the consistent predictions between Illustris and MassiveBlackII, this also implies good agreement between Illustris and the observations for both ends of the relation.

3.3.2 The M M Relation BH − DM Figure 3.7 shows the Illustris prediction of the M M relation. The best fit to the BH − DM simulation data is log(M /M⊙) = (1.27 0.02) log(M /M⊙) (7.72 0.24), with BH ± DM − ± R2 = 0.65, RMSE = 0.42. The overall scatter is higher than that of the M M . BH − ⋆,total In particular, the simulation over-produces galaxies with above-average BH masses for

12 MDM < 10 M⊙. The best fit to the simulated spiral sample is depicted by the dashed yellow line, which is very close to the one for all types (solid yellow line). Both the barred (green filled circles) and non-barred spirals (red crosses) follow the mean relation closely without falling systematically above or below the relation.

3.3.3 The M M Relation BH − halo Figure 3.8 compares the Illustris prediction of the M M relation with other BH − halo theoretical results and the observational constraints. Note that Mhalo was defined as the total mass of all particles (i.e., gas, dark matter, black holes and stars) bound to 76

Figure 3.6: The Illustris prediction of the M M relation is shown by the BH − ⋆,total solid yellow line, which represents the best fit to the simulated data: the slope and normalization are 1.51 0.02 and 8.58 0.23 (R2 = 0.74, RMSE= 0.37) for the ± − ± log(M /M⊙) log(M /M⊙) relation. While the red crosses indicate the barred BH − ⋆,total spiral sample from Illustris, the green filled circles indicate non-barred ones. The best-fit to the spiral sample is shown by the dashed yellow line. The observed M M BH − ⋆,total relations by Reines and Volonteri (2015) are represented by the solid blue line for local AGNs and the dashed blue line for E/S0 galaxies. These authors also suggested that spirals and disks follow the same relation with local AGNs. The relation by DeGraf et al. (2015) from the high-resolution cosmological simulation MassiveBlackII is represented by the magenta line. 77

Figure 3.7: The Illustris prediction of the M M relation is shown by the solid BH − DM yellow line, which represents the best fit to the simulated data: log(MBH /M⊙) = (1.27 2 ± 0.02) log(M /M⊙) (7.72 0.24) (R = 0.65, RMSE= 0.42). The dashed yellow ∗ DM − ± line represents the best fit to the spiral sample, which is shown by the red crosses (non-barred) and the filled green circles (barred). 78

Figure 3.8: The Illustris prediction of the M M relation is compared with BH − halo other theoretical predictions and the local observations. The solid cyan line is the best fit from Snyder et al. (2015) while the solid green line is the best fit from this work. The solid yellow line is the best fit to the spiral sample, which is shown by the red crosses (non-barred) and green filled circles (barred). The magenta lines are same as in Figure 5 of Ferrarese (2002): the solid line corresponds to the best fit using the prescription of Bullock et al. (2001) to relate vvir to vc, the dot-dashed line shows where the galaxies would lie if vvir = vc, and the dotted line shows where they would move to if vc/vvir = 1.8, as proposed by Seljak (2002). The solid blue line shows the observational determination of Bandara et al. (2009) while the dotted blue line shows the result from the simulation of Booth and Schaye (2010). The Illustris prediction of the M M relation is in remarkable agreement with other theoretical predictions BH − halo and the local observations. 79 the subhalo. Most self-regulating theoretical models of galactic physics predict a funda- mental connection between the central BH mass and the total mass of the host galaxy (e.g., Adams et al. 2001; El-Zant et al. 2003; Haehnelt et al. 1998; Monaco et al. 2000; Silk and Rees 1998). The galaxy models, which study the interaction between the DM halos of galaxies and baryonic matter, predict that halo properties determine the bulge and central BH masses (Cattaneo 2001; El-Zant et al. 2003; Hopkins et al. 2005). How- ever, there are inconsistencies in the observational studies: while some studies argued for (Seigar 2011; Volonteri et al. 2011), some argued against (Kormendy and Bender 2011) a coupling of the BH mass and the properties of the DM halo. It is not an easy task to investigate this relation observationally due to the uncertainities in measurement of the halo mass. Some of the first indirect observational evidence for the existence of a M M relation come from the study of Ferrarese (2002), which first derived BH − halo a correlation between σ⋆ and the galaxy’s circular velocity (vc) for a sample of 20 el- liptical galaxies and 16 spiral galaxies, then translated it to a M M relation. BH − halo While using the M σ relation from Ferrarese and Merritt (2000) to estimate the BH − ⋆ BH mass, the author demonstrated the effect of the method used to connect vc to Mhalo via v . The solid magenta line in Figure 3.8 shows the best-fit M M rela- vir BH − halo tion obtained by Ferrarese (2002) using the cosmological prescription of Bullock et al.

(2001) to relate vc and vvir. The dot-dashed magenta line shows the resulting relation if vvir/vc = 1 is assumed while the dotted magenta line presents the resulting relation if vvir/vc = 1.8 is used, as proposed by Seljak (2002). Among the relations obtained by Ferrarese (2002), the Illustris prediction agrees better with the solid magenta line, i.e., the one derived from the cosmological prescription of Bullock et al. (2001). The solid cyan line is the best fit from Snyder et al. (2015) while the solid green line is my best fit for Illustris. Note that Snyder et al. (2015) fitted the Illustris data by using or- thogonal distance regression method (Boggs and Rogers 1990) while a bisquare weights 80 was used here. Later, Bandara et al. (2009) investigated this relation by using gravita- tional lens modeling to determine the total mass. They estimated the BH masses from the M σ relation by G¨ultekin et al. (2009). They recovered a non-linear corre- BH − ⋆ lation between M and M , suggesting a slope of 1.67 that implies a merger- BH halo ∼ driven, feed-back regulated process for the growth of BHs. Based on their argument, the Snyder et al. (2015) fit with a slope of 1.69 implies a merger-driven, feed-back ∼ regulated process, and supports the conclusion of Sijacki et al. (2015) that both BH feedback and BH BH mergers are important ingredients to produce the M M − BH − ⋆,sph relation. Using self-consistent simulations of the co-evolution of the BH and galaxy populations, Booth and Schaye (2010) demonstrated a very good agreement between the observational determination of Bandara et al. (2009) and their simulation when the simulation was only tuned to match the normalization of the relations between MBH and the galaxy stellar properties. The authors also suggested a primary link to halo binding energy rather than the halo mass, with the central halo concentration playing a role in the relation’s scatter. It would be worth investigating this link with the future Illustris DM halo data catalogs (Chua et al. 2016) in a future study. In addition, the best fit to the spiral sample, shown by the solid yellow line in Figure 3.8, is consistent with the best fits to all galaxy types (solid cyan, green and magenta lines), implying that spirals in Illustris follow the mean relation very closely. This suggests the existence of a tight correlation between the galactic properties of star- forming, blue galaxies and their BHs. Also, no systematic behavior is observed based on bar morphology such as being well above or below the mean relation, suggesting the presence or absence of a bar does not affect the link between the BH and its host galaxy significantly. Moreover, the simulated galaxies present a smaller scatter around the mean M M relation when compared to the mean M M relation. It appears BH − ⋆,total BH − halo 12 that Illustris produces a wide range of BH mass at fixed MDM when MDM < 10 M⊙, 81 and that supports the result of Booth and Schaye (2010) about the importance of the central halo concentration.

3.4 Conclusions

This chapter investigates the BH galaxy connection beyond the bulge in the high- − resolution cosmological simulation Illustris. This work is complementary to Sijacki et al. (2015), which studied the black hole bulge scaling relations as a function of galaxy − morphology, color and specific star formation rate. Here these predictions are studied by comparing with other theoretical results and observational constraints at z = 0. First, the Illustris spiral arm morphology is studied in multiple wavebands and the existence of a 1:1 relation between the B- and K-band pitch angles in Illustris is presented, which is consistent with the observational findings of Seigar et al. (2006) and Davis et al. (2012). Then, the Illustris predictions for the BH mass, DM mass and halo mass were explored relative to pitch angle. I derived an Illustris prediction of the M P relation, that is consistent with observational constraints. More importantly, BH − I obtained a tighter correlation between pitch angle and DM/halo mass. Based on the Illustris predictions, it can be concluded that pitch angle can be a good tracer for DM/halo mass.

In addition, I focus on the z = 0 Illustris galaxies with 10.0 < log(M⋆,2R/M⊙) < 13.0 to study the M M , the M M and M M relations by comparing BH − ⋆,total BH − DM BH − halo with other theoretical studies and observational constraints. It was found that Illustris establishes very tight correlations between the BH mass and large-scale properties of the host galaxy, not only for early-type galaxies but also low-mass, blue and star-forming galaxies, regardless of bar morphology. The tight relations shown in this work are strongly suggestive that halo properties play an important role in determining those of the galaxy and its BH. 82 ,total ∗ ) M halo M ( log ) DM M ( is the mass of dark matter log -band pitch angle in degrees. DM K ) M ,total ∗ M ( log ) is the total mass of all particles (all types) bound BH , retrieved from the Illustris database. (7) M ( halo log M ustris subhalo ID. (2) Bar morphology: y for barred, n for 88 7.61 10.82 11.68 11.75 82 6.49 10.18 11.46 11.55 30 7.31 10.77 11.62 11.69 21 8.08 11.07 11.89 11.96 43 7.22 10.82 11.64 11.72 13 7.83 11.10 11.72 11.83 69 7.04 10.30 11.67 11.72 83 7.14 10.18 11.31 11.37 20 8.06 10.97 12.14 12.18 98 7.12 10.24 11.25 11.34 64 7.76 10.69 12.11 12.14 04 7.64 11.01 12.21 12.25 06 7.46 10.96 11.95 12.03 27 8.28 11.12 12.61 12.63 73 8.24 10.97 12.58 12.60 60 8.07 10.89 12.58 12.60 82 7.69 11.09 12.43 12.46 08 7.94 10.86 12.36 12.38 60 7.73 11.02 12.43 12.46 . (9) 17 7.84 11.06 12.23 12.27 ...... m the Illustris database. (8) . 1 3 2 4 2 2 1 2 2 3 2 3 3 3 1 0 2 4 3 3 K ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± -band pitch angle in degrees. (5) P R 24 84 58 14 43 72 74 29 39 79 71 54 62 55 30 88 05 35 53 49 ...... 53 27 19 22 59 14 74 23 75 24 04 20 42 30 17 10 38 23 26 19 38 20 11 20 47 12 22 14 77 10 25 16 35 18 82 21 79 16 81 9 ...... 3 2 4 1 2 3 4 3 2 1 4 1 1 3 0 2 1 2 1 1 R ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± P 44 34 65 67 72 32 06 24 00 96 76 18 41 51 76 67 08 32 60 22 ...... 12 9 44 22 65 22 06 14 64 21 65 22 89 22 43 26 78 13 38 27 00 19 65 16 24 16 39 14 99 10 88 10 92 16 88 19 45 24 82 9 ...... 1 3 1 3 3 3 3 3 3 3 1 3 1 1 2 1 1 1 1 2 B ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 68 72 70 61 82 37 53 53 74 15 18 89 87 94 56 86 52 40 01 25 ...... -band pitch angle in degrees. (4) B is the BH mass contained within the stellar half-mass radius BH 167869 n 12 66096 y 24 ID Bar41098 y P 15 (1) (2) (3) (4) (5) (6) (7) (8) (9) 183689 n 22 250637 n 19 278699 n 21 274690 y 21 287939 n 26 292278 n 13 293192 n 9 295983 n 28 308471 n 20 318403 n 18 315319 y 15 339972 n 14 348613 n 11 349838 n 10 353567 n 16 357033 n 18 360070 n 24 M bound to the subhalo, retrievedto from the the subhalo, Illustris retrieved database from the Illustris database. (6) is the total stellar mass bound to the subhalo, retrieved fro non-barred. (3) Table 3.1: Illustris Spiral Galaxy Sample. Columns: (1) Ill 83 ) halo M ( log ) DM M ( log ) ,total ∗ M ( log ) BH M ( log 63 7.53 10.78 12.34 12.38 04 8.26 10.98 12.43 12.44 76 7.93 10.88 12.37 12.39 11 7.79 10.94 12.34 12.37 40 7.74 10.79 12.19 12.23 24 7.51 10.86 12.20 12.24 60 7.50 10.82 12.15 12.18 39 7.84 10.90 12.24 12.27 55 8.15 10.80 12.28 12.30 17 7.60 10.70 12.21 12.24 28 6.94 10.56 11.99 12.05 78 7.61 10.82 12.18 12.21 83 6.51 10.16 11.88 11.91 52 7.78 10.94 12.19 12.21 33 7.88 10.91 12.17 12.21 13 7.62 10.79 12.15 12.18 77 7.69 10.91 12.17 12.20 76 7.55 10.77 12.14 12.18 45 7.18 10.75 11.85 11.91 14 7.19 10.71 12.00 12.06 18 7.28 10.84 12.02 12.08 04 7.56 10.77 12.04 12.07 47 7.63 10.86 12.05 12.09 83 7.39 10.91 12.00 12.05 47 8.05 11.09 12.36 12.39 38 7.79 10.77 12.14 12.17 ...... 1 3 2 1 2 2 2 2 3 1 1 1 2 1 1 4 2 2 3 2 3 3 1 2 2 3 K ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± P 31 96 71 80 15 88 22 95 50 57 71 14 29 98 43 83 02 23 86 16 35 01 04 83 75 41 ...... 46 20 31 15 67 16 45 17 74 19 26 25 74 20 82 10 89 13 10 18 57 9 79 28 68 19 42 30 70 16 67 13 07 15 51 20 62 24 59 18 31 15 88 23 28 20 43 21 05 17 69 8 ...... 1 3 2 1 2 2 1 1 2 2 2 2 1 4 2 1 4 2 2 3 2 3 2 1 2 1 R ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± P Illustris Spiral Galaxy Sample (continued) 43 14 24 31 62 96 38 35 31 50 75 10 07 06 23 98 82 73 20 23 81 19 46 53 96 42 ...... 60 20 32 17 90 18 21 18 32 18 58 21 41 22 23 11 82 15 11 18 16 21 28 27 47 21 43 26 38 17 46 13 03 19 72 20 51 19 13 18 66 21 68 28 14 17 73 22 36 20 89 9 ...... 1 3 2 2 2 3 1 1 2 3 2 3 1 3 2 1 3 1 2 3 3 3 3 2 1 1 B ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 67 92 15 92 07 07 11 95 52 09 44 10 52 53 78 09 34 68 89 02 86 73 75 36 52 58 ...... ID Bar365003 n P 19 (1) (2) (3) (4) (5) (6) (7) (8) (9) 373991 n 13 381137 n 19 376968 n 9 379786 y 15 377089 y 19 390346 n 19 393605 n 21 393984 n 11 396937 n 14 397204 n 19 397695 n 19 400083 n 26 401370 n 20 401856 n 26 401474 y 15 403323 n 14 403642 n 20 406286 n 20 407890 n 16 408954 n 20 412343 n 23 412515 n 27 417495 n 15 416531 y 24 417736 n 20 84 ) halo M ( log ) DM M ( log ) ,total ∗ M ( log ) BH M ( log 58 7.64 10.64 12.05 12.08 49 7.57 10.77 12.00 12.03 48 7.50 10.54 12.00 12.03 99 6.39 10.20 11.78 11.83 60 7.02 10.33 11.92 11.97 95 6.90 10.24 11.83 11.87 86 6.90 10.11 11.59 11.63 91 7.15 10.23 11.86 11.89 83 7.20 10.24 11.83 11.87 31 6.67 10.34 11.69 11.76 50 7.48 10.26 11.85 11.87 74 7.43 10.28 11.81 11.84 68 7.01 10.28 11.75 11.80 71 7.03 10.20 11.77 11.80 65 6.53 10.14 11.64 11.70 40 6.49 10.19 11.55 11.61 37 7.06 10.18 11.69 11.73 24 7.11 10.12 11.72 11.75 66 7.02 10.16 11.75 11.78 59 6.92 10.28 11.70 11.74 75 6.82 10.08 11.69 11.74 88 6.51 10.21 11.56 11.62 41 6.98 10.24 11.58 11.64 33 6.85 10.20 11.62 11.68 88 7.27 10.14 11.67 11.71 74 6.90 10.09 11.59 11.65 ...... 3 2 3 2 2 3 2 2 2 1 2 3 2 2 2 2 3 5 2 2 1 3 3 3 1 1 K ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± P 11 93 16 96 14 00 23 30 43 33 47 75 96 15 52 17 95 40 59 34 73 60 16 45 89 42 ...... 49 19 11 20 81 25 39 29 15 26 15 30 75 21 32 27 79 23 19 12 71 18 08 13 63 21 52 26 07 28 84 21 12 12 91 21 35 21 67 15 99 14 39 35 33 28 30 26 93 23 39 18 ...... 3 2 1 2 3 3 1 3 2 1 2 2 2 2 1 2 3 2 1 1 3 3 2 3 1 2 R ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± P Illustris Spiral Galaxy Sample (continued) 77 42 44 34 59 82 31 89 38 49 48 03 96 19 42 04 13 19 64 74 27 95 58 03 17 53 ...... 48 17 21 16 85 11 72 25 33 24 89 34 51 18 91 24 23 18 12 13 86 22 19 13 31 14 86 25 38 23 56 27 27 14 94 13 17 18 71 17 72 16 95 29 02 29 65 24 35 24 45 18 ...... 3 3 2 2 3 2 1 1 2 3 2 3 3 2 2 2 2 2 1 3 2 1 3 2 3 2 B ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 29 55 51 66 92 62 43 75 86 62 93 37 62 72 03 04 12 39 89 49 92 00 62 89 21 05 ...... ID Bar420546 n P 18 (1) (2) (3) (4) (5) (6) (7) (8) (9) 429333 n 16 432924 n 20 437661 n 25 437478 y 23 445191 n 34 453583 n 18 450907 y 26 450471 y 19 458148 n 16 461309 n 20 466545 n 10 461677 y 16 468379 n 23 468683 n 24 470338 n 25 475081 n 16 474410 y 15 474359 y 23 474123 y 18 475149 n 15 475122 y 38 478250 n 29 478741 n 24 479639 n 28 480862 n 19 85 ) halo M ( log ) DM M ( log ) ,total ∗ M ( log ) BH M ( log 40 7.07 10.23 11.64 11.68 71 6.88 10.13 11.60 11.64 80 6.22 10.11 11.50 11.58 92 6.80 10.15 11.48 11.54 63 6.84 10.09 11.50 11.56 46 6.87 10.12 11.62 11.66 93 6.10 10.05 11.48 11.56 72 6.10 10.14 11.50 11.58 95 6.26 10.53 11.36 11.44 75 7.06 10.15 11.61 11.64 57 7.06 10.13 11.63 11.66 32 7.21 10.11 11.61 11.64 25 6.63 10.08 11.55 11.60 97 6.71 10.15 11.52 11.58 76 6.95 10.12 11.53 11.56 39 6.66 10.14 11.46 11.53 84 6.65 10.12 11.47 11.53 22 6.22 10.08 11.45 11.51 89 6.13 10.16 11.37 11.46 66 6.60 10.16 11.39 11.46 51 6.97 10.17 11.36 11.44 85 6.71 10.14 11.33 11.40 45 6.36 10.14 11.26 11.34 ...... 2 2 3 3 2 2 2 2 3 2 2 3 2 2 2 2 1 3 2 2 2 2 2 K ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± P 66 50 92 50 63 35 11 41 20 35 47 72 33 65 20 59 24 74 69 23 73 94 26 ...... 44 19 26 22 71 28 26 29 19 22 92 15 06 24 54 29 97 30 11 19 81 18 35 19 29 24 54 24 54 24 47 23 80 26 11 31 20 33 14 32 50 27 29 26 36 28 ...... 2 3 3 3 2 1 2 2 2 2 1 1 2 2 2 2 1 3 3 2 2 2 4 R ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± P Illustris Spiral Galaxy Sample (continued) 26 85 40 53 11 56 22 33 20 69 55 58 04 08 93 02 06 48 64 06 30 04 21 ...... 20 22 88 18 82 27 38 28 77 30 53 17 85 23 18 29 72 21 65 23 36 17 19 13 77 22 80 27 60 24 60 22 69 24 01 29 40 26 99 29 91 28 65 25 68 29 ...... 2 2 3 3 3 1 2 3 2 2 1 2 1 2 2 3 2 3 4 1 2 2 2 B ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 03 36 34 76 60 42 95 43 13 17 64 96 63 69 98 29 07 98 60 40 32 58 95 ...... ID Bar484500 n P 20 (1) (2) (3) (4) (5) (6) (7) (8) (9) 484893 n 19 487887 n 30 487035 y 29 489973 n 29 488826 y 17 488002 y 29 491764 n 23 492528 n 20 493024 n 26 493332 n 18 494358 n 13 494540 n 17 494646 n 26 498275 n 23 497326 y 20 501185 n 20 506252 n 25 503382 y 28 508079 n 28 510148 n 29 518358 n 23 522537 n 31 Chapter 4

The Ring Formation Mechanisms

There are different types of galaxy rings such as polar (e.g., A 0136-0801, Whitmore et al. 1990), collisional (e.g., Cartwheel Galaxy), accretion (e.g., NGC 7742, de Zeeuw et al. 2002), and the more familiar rings formed due to secular evolution (e.g., NGC 1291, Buta and Combes 1996). These latter rings are commonly found in barred disk galaxies, though they are seen in non-barred disk galaxies as well (Grouchy et al. 2010). The origin of rings is generally related to either slow secular evolution or environ- mental processes. Buta and Combes (1996) pointed out the possibility that non-barred ringed galaxies may include both internally and externally generated rings. Hence, non- barred ringed galaxies are among the ideal objects to study the role of both the internal dynamics of galaxies and the physics of accretion/interaction mechanisms. This chapter reviews possible scenarios of ring formation, focusing on the aspects most relevant to the rest of the dissertation: the origin of Hoag-type galaxies.

86 87 4.1 Resonance Rings

Rings formed due to secular evolution, i.e., resonance rings, are distinguished by the radii at which they occur: nuclear rings, inner rings, and outer rings (see Figure 4.1). Inner rings lie in the intermediate zone between the bulge and the spiral arms or disk, generally near the ends of a bar, if present. While outer rings have the largest relative radius, nuclear rings have the smallest of the three. The ring classification is very straightforward for barred galaxies. The bar usually fills an inner ring in one dimension so that other types can easily be distinguished; however, the absence of the bar creates confusion for the ring type in non-barred galaxies (Buta and Combes 1996).

Figure 4.1: Resonance rings are distinguished by the radii at which they occur (Buta and Combes 1996): nuclear rings (left: NGC 6782; image taken from the Hubble Legacy Archive), inner rings (middle: NGC 1433; image taken from the NED), and outer rings (right: NGC 1543; image taken from the NED).

A successful theory for ring formation is that rings are formed by gas accumulation at the Lindblad or corotation resonances in response to gravitational torques from the bar pattern (e.g., Sellwood and Wilkinson 1993; Buta and Combes 1996). In this picture, a mild oval distortion in the disk forces spiral structure in the gas (Sanders 1977), and the evolution of such structures creates tangential forces on the gas. The gas acquires 88 angular momentum and is driven outwards between the corotation (CR) and the outer Lindblad resonance (OLR), and loses angular momentum between the CR and the inner Lindblad resonance (ILR) (Combes 1988). While outer rings form at the OLR of the bar, inner rings occur at the inner second harmonic resonance of the bar, and nuclear rings lie at the ILR of the bar (e.g., Combes and Gerin 1985; Buta 1986). The simulations by Schwarz (1981, 1984) demonstrated that even a small oval distortion in the disk could form an outer ring at the OLR within a few Gyr. This time scale for ring formation can be longer for a weaker bar. In addition to the strength of an oval potential, Athanassoula and Bosma (1985) pointed out the importance of the amount of the gas, which was left in the region outside the CR after formation of the bar in the disk. This may explain the absence of the ring structure in some strongly-barred galaxies. Although this bar-driven ring theory is widely accepted, it does not offer convincing reasons for the existence of rings in non-barred galaxies. Rings are less common in non- barred galaxies, but they are by no means rare (Buta 1991). Sellwood and Wilkinson (1993) argued that the presence of rings in non-barred galaxies might be explained by bars being unseen in the optical or that they have dissolved, leaving the rings that formed when the bar was stronger. In fact, it is shown that some bars can be only detectable in the near-infrared where the galaxy is more transparent to dust (Zaritsky and Lo 1986; Rix and Rieke 1993). Near-infrared light efficiently penetrates the dust and is more sensitive to the older stellar population, which is typically found in bars. However, blue light is a good passband for detecting rings, dust, and young stars, hence bars are less prominent in such a waveband. Therefore, some ringed galaxies that are classified as non-barred in optical may reveal a bar structure in the near-infrared (e.g., NGC 3147, Casasola et al. 2008). In the case of bar dissolution, the bar might have been destroyed through either tidal interaction with a companion, or through gas accretion 89 and mass concentration, but these bar destruction mechanisms raise some questions about the survival of rings (Buta and Combes 1996). On the other hand, Athanassoula (1996) examined how bar dissolution affects a simulated outer ring and found that a ring survives for a long time after the bar dissolves. It is also possible that the presence of a steady spiral density wave can create gravity torques on the gas similar to the bar, and this could explain the existence of some rings in non-barred galaxies (e.g., Mark 1974). Buta and Combes (1996) argued that this mechanism is not sufficient since spirals are not steady enough in the absence of bars. However, Rautiainen and Salo (2000) demonstrated that in models with a hot disk that never formed a bar, a spiral potential can still effectively form a ring at the spiral ILR.

4.2 Galaxy Galaxy Interactions − Galaxy-galaxy interaction theories have been proposed to explain the presence of rings in non-barred galaxies. One such theory suggests that a ring could be the response to the steady tidal action of a bound companion since the non-axisymmetric pertur- bation is very similar to that of a bar (Combes 1988; Helmi et al. 2003). A violent encounter can also be responsible for ring formation as in the case of collisional ring galaxies (Lynds and Toomre 1976; Appleton and Struck-Marcell 1996). The prototype of this class is the Cartwheel galaxy (see Figure 4.2). In a collisional ring scenario, head-on collisions between disk galaxies create circular density-waves that propagate outwards from the center and induce subsequent starbursts that form a luminous ring. This is well-described by Athanassoula and Bosma (1985) and numerically simulated by Mapelli et al. (2008). Milder interactions, such as accretion, are also proposed for ring formation as in the case of accretion rings (see the left panel in Figure 4.3). They are thought to be made of material from a gas-rich companion (Schweizer et al. 1987, 1989; de Zeeuw et al. 2002). 90

Figure 4.2: The prototype of collisional ring galaxies is the Carthwheel Galaxy. Image taken from the Hubble Legacy Archive. 91

Figure 4.3: Left: an example for accretion rings: NGC 7742 (de Zeeuw et al. 2002). Right: the prototype of polar ring galaxies: NGC 4650A (e.g., Arnaboldi et al. 1997). Images taken from the Hubble Legacy Archive. 92 There are also polar ring galaxies (PRGs), which have a polar ring or disk which or- bits in a plane nearly perpendicular to the central galaxy (Schweizer et al. 1983; Sparke 1986; Whitmore et al. 1987, 1990; Rubin 1994). While roughly 0.5% of all nearby S0 galaxies appear to have polar structures, the percentage increases to 5% when cor- ∼ rected for selection effects, e.g., non-optimal viewing angle, very low surface brightness, limited lifetime of the ring (Whitmore et al. 1990). The right panel in Figure 4.3 shows the prototype of PRGs, NGC 4650A, which has a polar disk structure (Arnaboldi et al. 1997; Iodice et al. 2002; Gallagher et al. 2002; Swaters and Rubin 2003). The currently favored formation scenarios for PRGs are: (1) a major dissipative merger, where a PRGs results from a polar merger of two disk galaxies with unequal mass (Bekki 1998; Bournaud and Combes 2003; Bournaud et al. 2005); (2) a scenario by which the ring is formed from material accreted by a lenticular or elliptical galaxy from another galaxy (Bournaud and Combes 2003; Knapen et al. 2004; Sil’chenko and Moiseev 2006); (3) infall from filaments, where a polar disk may form if gas falls into a galaxy along cosmic filaments that are inclined to the stellar disk (Macci`oet al. 2006; Brook et al. 2008). The tidal accretion-induced polar ring is formed more readily than the merger-induced sce- nario (Bournaud and Combes 2003; Knapen et al. 2004; Sil’chenko and Moiseev 2006). For the infall from filament scenario, Snaith et al. (2012) found stringent constraints on the structure of the polar disk: a sub-solar metallicity Z = 0.2Z⊙, a flat metallicity gradient, and a younger age (< 1 Gyr) with respect to that of host galaxy (4-5 Gyr).

4.3 Hoag-type Galaxies

Hoag’s Object (PGC 054559; Hoag 1950) is a nontypical ringed galaxy with its peculiar morphology: an elliptical-like core with a nearly circular outer ring, and no signs of bar or stellar disk. Hoag-type galaxies, which have a fair resemblance to Hoag’s Object, are extremely rare (Schweizer et al. 1987) and their origin is still under debate. Brosch 93 (1985) proposed the ring of Hoag’s Object was formed in a similar manner to rings in barred galaxies (i.e., the bar instability scenario involving slow internal motions). However, Schweizer et al. (1987) showed that the inner core is a true spheroid, not a disk, and suggested a major accretion event where a spheroid galaxy accreted and gas was transferred from a colliding and/or a small companion galaxy into a ring. They attributed the absence of a merging signature to aging (2-3 Gyrs) of the accretion event. Both Brosch (1985) and Schweizer et al. (1987) ruled out the possibility of Hoag’s Object being a classical collisional ring galaxy seen face-on. As Brosch (1985) pointed out, this scenario requires a high-level tuning of parameters such that a companion would have to pass through very close to the center of the disk and almost perpendicular to the plane with its relative velocity closely aligned into the line of sight. If all of these conditions were not met, the resulting galaxy would be a non-uniform or elliptical ring with an off-centered central object, so he narrowed down the probability of forming Hoag’s Object via this mechanism to 10−5. Furthermore, Schweizer et al. (1987) discussed that, had Hoag’s Object been formed by this mechanism, it would have implied a relative velocity of order 100 km s−1 of the two galaxies and yet the relative velocity between the core and the ring is almost zero. Whitmore et al. (1990) considered this galaxy as a system related to PRGs, emphasizing that Hoag-type galaxies are unlikely PRGs, but they may have similar evolutionary histories since they share some characteristics with them. A recent study by Finkelman et al. (2011) considered accretion of gas from the intergalactic medium (IGM) on a pre-existing elliptical galaxy to explain both the peculiar structure and the kinematics of Hoag’s Object. Wakamatsu (1990) examined a list of Hoag-type galaxies and showed that they often show elongated cores with disk-like features (e.g., NGC 6028), and probably represent an evolved phase of barred galaxies where the bar has been almost completely destroyed. However, Finkelman and Brosch (2011) focused on UGC 4599 as the nearest Hoag-type 94

Figure 4.4: Hoag’s Object (PGC 054559; Hoag 1950). Image taken from the NED. 95 ringed galaxy. Having failed to detect a bar or a central disky component, the au- thors rejected the hypothesis that UGC 4599 is a barred early-type galaxy where the bar dissolved over time, and suggested that the cold accretion of gas from the IGM can account for the observed peculiar properties of the galaxy, although Macci`oet al. (2006) demonstrated that gas accretion on a spheroidal is more likely to produce a S0 galaxy, rather than an elliptical-like galaxy. While Moiseev et al. (2011) considered UGC 4599 as a possible face-on PRG, Finkelman and Brosch (2011) emphasized that face-on polar rings with a Hoag-like structure were rarely discovered, despite a dedi- cated search (see Taniguchi et al. 1986; Moiseev et al. 2011), implying that ellipticals with round bulges and almost circular rings such as Hoag’s Object are exceedingly rare. Finkelman and Brosch (2011) also did not rule out the possibility that some of the gas in the ring was accreted from a close companion during the evolution of the galaxy. Qualitatively, all these described scenarios are capable of describing a number of pe- culiarities of Hoag-type galaxies; however, the small number of known objects do not provide definite conclusions about their nature, evolution, and systematic properties. Chapter 5

A photometric study of the peculiar ringed galaxy: PGC 1000714

A slightly modified version of this chapter has been published in Monthly Notices of the Royal Astronomical Society with the following bibliographic reference: B. Mutlu-Pakdil, M. Mangedarage, M. S. Seigar and P. Treuthardt. A photometric study of the peculiar and potentially double ringed, non-barred galaxy: PGC 1000714. MNRAS, 466:355-368, April 2017.

5.1 Overview

This chapter presents a photometric study of PGC 1000714 (2MASX J11231643 0840067), which appears to be a genuine ring galaxy. While it was − included in a number of catalogs, it has not yet described in the literature. Although it is classified as (R)SAa in the Third Reference Catalogue of Bright Galaxies (RC3,

96 97 de Vaucouleurs et al. 1991), this galaxy has a fair resemblance to Hoag’s Object in the optical and near-ultraviolet (NUV) bands, and that makes it a good target for a detailed study of Hoag-type galaxy structure. Since Hoag-type galaxies are extremely rare and peculiar systems, it is necessary to increase the sample of known objects by performing detailed studies on other possible candidates to derive conclusions about their nature, evolution, and systematic properties. Such peculiar systems help our understanding of galaxy formation in general, since they represent extreme cases, providing clues on formation mechanisms. Since the photometric properties retain enough memory of the evolutionary processes that shaped the galaxies, one can retrieve valuable information about the nature of the structural components via a photometric study. Therefore, the surface photometry of the galaxy was performed by using both archival data (in NUV and infrared) and new data acquired in optical wavebands (BVRI ). By reconstructing the observed spectral energy distribution (SED) for the central body and the ring, the stellar population properties of the galaxy components were recovered. This work has revealed, for the first time, an elliptical galaxy with two fairly round rings. The central body follows well a r1/4 light profile, with no sign of a bar or stellar disk. The results suggests different formation histories for the galaxy components. Possible origins of the galaxy are discussed, and it is concluded that a recent accretion event is the most plausible scenario that accounts for the observational characteristic of PGC 1000714. This chapter is organized as follows. Section 5.2 describes the general properties of the galaxy. After a description of the imaging data in Section 5.3, the methodology and detailed results of isophotal fitting are presented in Section 5.4. To further enhance the visibility of the outer ring, a two-dimentional (2D) image analysis was performed and the results of the photometric decomposition are presented in Section 5.5. After focusing on the rings in Section 5.6, the stellar population properties are examined in Section 5.7. Then, the implications of the observational evidence are then discussed in 98 Section 5.8 and finally the main results are summarized in the conclusions (Section 5.9).

Throughout this chapter, unless specified otherwise, a standard cosmology of H0 =

−1 −1 73 km s Mpc ,ΩΛ = 0.73 and ΩM = 0.27 is assumed and magnitudes are used in the Vega system.

5.2 Observational Data

The NUV imaging analysed here was obtained from archival GR6/GR7 data release of the GALEX All-Sky Imaging Survey. The background subtracted intensity map image corrected for the relative response was used. The image covers a field of 2.5 arcmin radius on the sky with a scale of 1.5 arcsec pixel−1. The resolution FWHM is 5.3 arcsec. Due to the large FWHM, the GALEX image was not very suitable for a detailed study of the inner part of the galaxy. Therefore, this data was used to investigate the outer ring only. The optical images were observed at the 2.5-m du Pont telescope at the Las Cam- panas Observatory on 19 January 2006 with the Direct CCD Camera during dark time. The Direct CCD Camera is a 2048 2048 pixel camera, with a pixel scale of 0.259 × arcsec pixel−1 resulting in a 530 arcsec 530 arcsec field of view. The image of PGC × 1000714 just happened to appear in the background of another galaxy being observed, off to the bottom right of the field of the camera. This resulted in part of the galaxy in question falling off the side of the chip. Total exposure times were 2 360 s in the × B band, 2 180 s at V, 2 120 s at R, and 2 180 s at I. The seeing measured in this × × × particular set of images were 1.17, 1.11, 1.01, and 0.96 arcsec in the B, V, R, and I bands, respectively. For each filter, 20 bias frames were taken and combined using the ZEROCOMBINE task in IRAF. The CCDPROC command was then used to remove the bias level. The images were flatfielded using twilight flats. Several flat-field exposures were taken and 99

Figure 5.1: This figure illustrates the sky estimation method adopted here. Left: B- band image of PGC 1000714, overlayed by the ellipses that were used to estimate the sky value. The length of the semimajor axis (SMA) for the inner and outer ellipse is 42 arcsec and 78 arcsec, respectively. The red regions are masked areas. Right: The isophotal intensity in counts of fixed ellipses fitted every 10 pixels. The region between the vertical dashed lines was used to estimate the sky value, which is indicated as horizontal dashed line together with the two dotted lines at 1σ. ± combined using the FLATCOMBINE task, and CCDPROC was then run to flatten the science images. Finally, the images were combined using the IMCOMBINE task by taking an average of the images and rejecting cosmic rays. A mask file was created that contained positions of bad pixels and columns and the IMREPLACE task was used to account for these. Finally, the IMTRANSPOSE command was used to correct the orientation of the chip, so that North is up. The images were taken during photometric conditions. Standard stars from the Landolt Equatorial Standard Stars Catalog (Landolt 1992) were observed at different airmasses through the night in all four filters. It was found that an aperture size of about 4 FWHM worked best for performing aperture photometry of the standard star × images. The photometric calibration was then performed using the DAOPHOT package 100 within IRAF. Due to the high resolution and the small seeing FWHM, this work mostly relies on the optical data. To determine the sky level, an ellipse fitting routine was applied to the original image in linear steps of 10 pixels ( 2.6 arcsec) between successive ellipses. The program was ∼ forced to extend the fit well beyond the galaxy with the fixed ellipticity (ǫ) and position angle (P.A.) of the outer ring. By checking the counts at these ellipses as a function of radius, it was found that the ellipse at 40 arcsec leaves the galaxy and enters the background region, where the flux is constant. After visual confirmation that these ellipses were really outside the visible galaxy, the mean and standard deviation of the fluxes between 42 and 78 arcsec were used as the final sky values (see Figure 5.1). This method is very similar to the one presented by Pohlen and Trujillo (2006). As the surface photometry is sensitive to the adopted sky value, the sky values were checked in a few ways. First, the mean of 5 median fluxes that were obtained from 10 ∼ 10 pixel boxes that was positioned near the galaxy-free corners of each image was × calculated. Second, the IRAF DAOEDIT task was used to estimate sky and sky-sigma of several stars. Finally, the most common pixel value was checked by using the mode in IMSTAT task. Results from these different methods agree with the adopted values within the 2σ uncertainty, which means the adopted sky values and their uncertainties are well-defined. The near-infrared (NIR) images came from the Two Micron All Sky Survey (2MASS), which covers the entire sky in the filters J (1.25µm), H (1.65µm) and Ks (2.17µm). The images of a square field of view of 4 4 arcmin2 on the sky with a scale of 1 arcsec × pixel−1 were used. The seeing FWHM of the J, H, Ks-band images were 2.48, 2.39, and 2.54 arcsec, respectively. The calculations here rely on the photometric zero point given in the keyword MAGZP in the header of each image, specifically 20.81, 20.41, and 19.92 mag. The sky values were simply taken from the keyword SKYVAL in the FITS 101 header of each image. Unfortunately, the 2MASS survey has a weak sensitivity to low surface brightness due to the high brightness of the night sky in the NIR range and short exposures. For this reason, the part of the galaxy beyond the half light radius are

−2 unseen beyond the isophotes fainter than µK =20 mag arcsec . The mid-infrared (MIR) images come from the Wide-Field Infrared Survey Explorer (WISE), which covers the sky in four bands centered at 3.4µm (W1), 4.6µm (W2), 12µm (W3), and 22µm (W4) with an angular resolution of 6.1, 6.4, 6.5, and 12.0 arcsec, respectively. The intensity images produced by match-filtering and coadding multiple 7.7 s (W1 and W2) and 8.8 s (W3 and W4) single-exposure images were used. The pixel scale of the images was 1.375 arcsec pixel−1. The sky values were estimated with the same method used in the optical bands.

5.3 General Properties

The images of PGC 1000714 in different filters are shown in Figure 5.2. The appearance of the galaxy in the NUV and optical wavebands resembles that of an almost round elliptical object with a detached outer ring, fairly similar to Hoag’s Object. The outer ring is more clearly visible in the NUV and B-band images. While still visible in the I band, it disappears in the J band. Both the infrared and optical images clearly show that the central body is the dominant luminous component but the central body disappears in the W4 image. The galaxy was investigated by deriving the high-frequency residual image, which was produced by taking the ratio of the original image to its median filtered image, which was computed with the FMEDIAN package in IRAF, where each original pixel value was replaced with the median value in a rectangular window. Using a small window size highlights the prominent galaxy substructure while using a larger window size highlights the prominent large-scale galaxy components. First, a window size of 5 102

Figure 5.2: The images of PGC 1000714 in the NUV,B, I, J and W3-band are shown from left to right, respectively. For each image, the size is 75 90 arcsec, and the north × is up while the east is on the left. The appearance of the galaxy in the NUV and optical bands is fairly similar to Hoag’s Object.

Figure 5.3: The high frequency residual image for PGC 1000714 is shown in the B- band (left panel) and in the I band (right panel). These images are obtained using a window size of 5 5 pixels in order to get the optimum enhancement of the galaxy inner × structures. Any substructure is expected to appear as a region with a value greater than 1. The darker regions are characterized by a value greater than 1, and lighter regions by a value less than 1. The B and I - band residual images do not show any underlying structure. For both image, the size is 75 90 arcsec, and the north is up while the east × is on the left. 103

Figure 5.4: Left-hand panel shows the high-frequency residual B-band image of PGC 1000714, which is obtained using a window size of 301 301 pixels in order to get the × optimum enhancement of the galaxy prominent structures. Any prominent component appears as a region with a value greater than 1. The bright central object appears as star-like (or bulge-like). The length of the semimajor axis (SMA) for the yellow ellipse is 18 arcsec, while those of the white ones are 25 and 41 arcsec. All three ellipses share the same center. Right-hand panel shows the residual image that is produced by fitting ellipses to the core and subtracting the BMODEL image from B-band image. Two background sources can be seen in the region where the bulge was subtracted, within the outer ring. The core does not show any underlying structure. 104 5 pixels was used in the optical bands, failing to detect any prominent substructure × (see Figure 5.3). Then, a window size of 301 301 pixels was used in the B band, and × this highlights the central body and the outer ring (see Figure 5.4). Their extensions are roughly estimated by ’eye’. The outer ring has an orientation and ellipticity similar ◦ ◦ to the core, with slightly higher values: P.A.ring = 60 , P.A.core = 50 , ǫring = 0.20,

ǫcore = 0.15. Note that, in Figure 5.1, the central body extends through the outer ring which appears as a bump starting at 20 arcsec and fades away at 40 arcsec. The general information reported by the NED is listed in Table 5.1. The redshift of PGC 1000714 is z = 0.0257 0.0001 (v = 7717 38kms−1; Mahdavi and Geller ± r ± 2004). The number of neighbors of PGC 1000714 is checked by searching the NED for nearby galaxies with known and similar . A similar search was done by Finkelman and Brosch (2011) to identify the environment of the galaxy in question. The search is made within a radius of 120 arcmin of the object and a redshift between

−1 vr = 6700 and vr = 8700 km s . It is found that PGC 1000714 does not lie within the boundary of a recognized cluster of galaxies and its nearest neighbor GALEXASC J112227.40 081817.0 is about 24.95 arcmin away. Therefore, its neighborhood is char- − acterized as a low-density environment.

Table 5.1: General information for PGC 1000714 from the NED

PGC 1000714 RA(J2000) 11h23m16.44s Dec.(J2000) 08d40m06.5s − − (v ; km s 1) 7717 38 r ± Redshift (z) 0.025741 Distance (Mpc) 110 105 5.4 Isophotal Analysis

5.4.1 Ellipse Fitting

Ellipse fitting to isophotes of galaxies is a common practice to obtain semi-major axis (SMA) profiles for surface brightness, ellipticity (ǫ; defined as 1 b , where a and b − a are, respectively, the semi-major and semi-minor axes of the best-fitting ellipse), and position angle (P.A.; measured counterclockwise from the +y-axis). For a smooth and featureless spheroidal galaxy the projected ellipses are concentric with the same P.A. and ǫ. The intensity fluctuates along the best-fitting ellipse and can be quantified by the following Fourier expansion: ∞ I(θ)= I0 + X(An sin(nθ)+ Bn cos(nθ)) (5.1) n=1 where I0 is the intensity of the best-fitting ellipse. The relative radial deviation of the isophote from elliptical shape is described by the coefficients sn and and cn, which are related to An and Bn according to

An Bn sn = dI , cn = dI r r | dr | | dr | where r = a√1 ǫ with ǫ the ellipticity of the best-fitting ellipse − (Milvang-Jensen and Jørgensen 1999, and references therein). An ellipse can be described by the first and second order Fourier coefficient, there- fore the Fourier expansion along the best-fitting ellipses gives zero values for A0, A1,

A2, B1, B2. The deviations from pure ellipses are then expressed by the higher- order coefficients An and Bn with n > 3. In particular, the B4 coefficient is im- portant because the dominant harmonic mode in most cases is the fourth order co- sine function and it defines the c4, which is an indicator of a disky (B4 > 0) or a boxy (B4 < 0) structure (Milvang-Jensen and Jørgensen 1999). Edge-on bars pro- duce boxy isophotes (e.g., Kuijken and Merrifield 1995; Bureau and Freeman 1999; 106 Bureau and Athanassoula 2005) while face-on bars may produce disky isophotes with the superposition of a prominent bulge (e.g., Athanassoula et al. 1990; de Souza et al.

2004; Gadotti and de Souza 2006). Although the B4 is dominant, one should not ignore the contribution of the A4. The A4 is indeed negligible when the second and fourth order components have roughly the same position angle because the fitted ellipses will be aligned with the bar in the bar region (Gadotti et al. 2007). However, Gadotti et al.

(2007) showed that the contribution of the A4 can be as large as half of the B4 and suggested that the significant contribution of the A4 may indicate the presence of con- spicious spiral arms or rings, which might contribute to the second and fourth order components differently than the bar. The isophotal structure and luminosity profile of the galaxy are analyzed by using the IRAF/STSDAS task ELLIPSE, which follows the procedure outlined by Jedrzejewski (1987), that fits a set of elliptical isophotes to an image. Before the ellipse fitting, all foreground stars and other contaminants were identified and masked by using SExtractor (Bertin and Arnouts 1996). In ELLIPSE, ǫ and P.A. were allowed to vary. To look for a possible varying center, the center was also left as a free parameter during ellipse fitting. However, the typical shifts obtained were less than a pixel (< 0.26 arcsec). In order to search for hidden structures in the central core, the fitting was restricted to the radius of r = 18 arcsec to avoid light contamination by the outer ring (see Figure 5.4, the yellow ellipse). The BMODEL task was then used to construct the core model. Subtraction of the model revealed that there are two background/foreground sources located between the radius of 5 and 10 arcsec: one is in the north-west, and other ∼ ∼ is in the south of the center. These sources are masked manually, and the ellipse fitting was repeated. This time, subtraction of the model from the actual image resulted in a very clean image (see the right-panel in Figure 5.4). The residual image shows no sign of an underlying substructure in the core. Although the core is modelled quite well, 107 there is some asymmetry that is only visible on the scale of the point spread function (PSF), and this may be completely caused by the seeing effect. The geometric properties derived from the optical, W1 and W2 images are shown in Figure 5.5. The results from the optical data are remarkably in good agreement. The ǫ and P.A. profiles display a clear systematic behaviour and follow a well-defined pattern. The central body is nearly round with an ellipticity value of ǫ 0.1 0.05 with the ∼ ± major axis at P.A.= 50◦ 10◦. The isophotal shape parameters, i.e., c and s , stay very ± 4 4 close to zero for the inner roughly 5 arcsec, as expected for an elliptical or round object.

The B-band c4 profile does not experience any significant variation, implying the core is featureless at shorter wavelengths. At longer wavelengths, the c4 values become slightly higher, which might suggest an old underlying substructure. However, the absence of a systematic behaviour in the optical c4 profiles (such as becoming negative or positive) is consistent with the absence of a bar-like feature. Although the c4 profile becomes positive in the W1 and W2, this can be the result of the seeing effect instead of a hidden structure. The resolution FWHM in the MIR data is much larger than the seeing in the optical data. The seeing or resolution disk has severe influence within the radius of 2 FWHM (Graham 2001; Huang et al. 2013). For the MIR data, the 2 FWHM × × roughly corresponds to 12 arcsec which covers most of data points. Therefore, one cannot draw any firm conclusion from the MIR data. The BVRI -band images have the largest ratio (Re/FWHM; see Table 5.2 for Re), hence the derived properties are more reliable when compared to those derived from the MIR images. In any case, the profiles are not very reliable near the center. This is because the inner light gets distributed to larger radii due to the PSF. This can be the reason of the peak in the ǫ which is observed just outside the FWHM of the PSF, in the BVRI -bands. Similar to the c4, the s4 values within the radius of 5 arcsec are all consistent with zero value. Outside this inner region, the s4 values get as high as that of the c4. As suggested by Gadotti et al. 108

(2007), these higher s4 values might be caused by a spiral or a ring.

Figure 5.5: Left-top panel shows the variation of ellipticity. Only the data whose error is less than 0.04 in ǫ is displayed. Left-bottom panel shows the variation of position angle (P.A.). Only the data whose error is less than 20◦ in P.A. is displayed. Both profiles are fairly smooth. Right-top panel shows the variation of c4. Only the data whose error is less than 0.02 in c4 is displayed. Right-Bottom panel shows the variation of s4. Only the data whose error is less than 0.02 in s4 is displayed. Both c4 and s4 profiles are close to zero within inner 5 arcsec, then they start to experience some higher variations toward the outer radii. The FHWM of the PSF for each band is shown with the red vertical dotted line.

Based on the optical data, one can conclude the central body is truly not barred. The core is further investigated with 2D image decompositions to check if the relatively higher values of the c4 and the s4 at the longer wavelengths are caused by an underlying substructure (see Section 5.5). 109 5.4.2 Radial Profile

The color profiles were derived from the ellipse fits in each band separately. The NUV profile is noisy compared to other profiles. Both the NUV and infrared colors suffered significantly from the large seeing effect within 2 FWHM. Only the optical colors were × well estimated. A careful inspection of the optical profiles revealed a fairly constant color index (see Figure 5.6). It was found that the optical color index at the inner region ( 1-12 arcsec) is fairly constant, but starts to vary slightly outside 12 arcsec. ∼ ∼ To examine the light distribution within the core, the core profiles in the BVRI bands were fitted with a S´ersic r1/n profile (Sersic 1968) with the least-squares minimization IDL package MPFIT. The optical images were used due to the smallest seeing effect. The S´ersic model converges toward the de Vaucouleurs law, yielding a value for n that is quite close to 4. To look for a possible disk component, I also tried to fit with the S´ersic bulge + exponential disk profile. However, this model could not fit the data, indicating the absence of a stellar disk structure. The best-fit in the B band indicates that the single r1/4 component matches the light profile well over a wide 6 mag range. Figure ∼ 5.7 shows the surface brightness profile of the core with a S´ersic profile fit in each optical band as a function of radius. The lower panel shows the residuals from the best-fitting model profile to highlight deviations from the model. While the B-band residuals are consistent with the zero value, the residuals toward the longer wavebands hint toward the presence of a reddish substructure. These residuals are further investigated with 2D image decompositions in Section 5.5.

5.5 2D Image Analysis

After extracting the 1D profiles, the galaxy was further investigated by deriving a 2D B-I color index map (see Figure 5.8). To correct for the PSF differences in each band, 110

Figure 5.6: The color gradients are shown: the NUV B, optical (B V , B R, B I), − − − − NIR (I J, I H, and I Ks), and MIR (I W1, I W2, and I W3) color plots. − − − − − − In the JHKs bands, the outer parts of the galaxy are unseen beyond 8 arcsec. The ∼ vertical red dashed line indicates the FWHM of the PSF in each band, e.g., top-left: 5 arcsec for NUV, 1 arcsec for optical, 2.5 arcsec for NIR, and 6 arcsec for ∼ ∼ ∼ ∼ MIR. The optical color indices of the galaxy present fairly constant radial profile, the others are not reliable due to the larger FWHM. 111

Figure 5.7: The best S´ersic model fit for the BVRI -bands. The best fit r1/n model in the B band gives n = 4, and the best fits in other bands give n values very close to 4. Re is the effective half light radius of the galaxy. The lower panels show the residuals from the best-fit S´ersic profile to highlight deviations from the model. 112

Figure 5.8: The 2D B I color index map is shown. Note that the light green color − scale reveals the existence of a diffuse inner ring. Both the inner and outer ring are pointed by the white arrows. The size is 79 110 arcsec, and the north is up while the × east is on the left. 113

Figure 5.9: The 2D image decomposition is shown for the BVRI bands (from left to right respectively). For each image, the size is 66 117 arcsec, and the north is up × while the east is on the left. Top panel displays images of the original data, middle panel displays the residuals from S´ersic bulge model generated by GALFIT, in which the initial parameters are allowed to vary. Bottom panel displays the residuals from the S´ersic bulge model, but this time the initial parameters are fixed to the 1D fitting results obtain from MPFIT. 114 IRAF task GAUSS was used to convolve the B-band image with the PSF in the I -band, and convolve the I -band image with the PSF in the B band. Then, the color map was obtained by transforming the images from intensity units into magnitude units, and correcting for the corresponding zero-point magnitudes, and then simply subtracting these two images. The 2D map reveals the existence of a diffuse inner ring with no sign of a bar. The existence of the inner ring explains the characteristic residuals seen in

Figure 5.7. In addition, it is consistent with the c4 and the s4 profiles that are derived in Section 5.4.1. The 2D image decomposition was performed by using the GALFIT profile fitting code (Peng et al. 2010). The outer and inner ring were carefully masked to prevent an overestimation of the outer parts of the galaxy. Then, the galaxy was modelled by using the same two functions (S´ersic function and S´ersic bulge + exponential disk function), while taking the PSF into account. Since GALFIT requires initial guess parameters (µe,

Re, n) to fit the data, the results from the 1D fitting were exploited to provide suitable initial guesses. These parameters were first allowed to vary, and then fixed to be able to compare the residuals. Similar to the previous fitting, the best fit of S´ersic function yields a value for n that is quite close to 4 when the input parameters are allowed to vary. The S´ersic bulge + exponential disk function does not fit the data. The residual images generated by GALFIT are shown for in Figure 5.9 for the optical bands. Table 5.2 lists the fitting results from GALFIT for two cases (fixed and free parameters), which are quite consistent with each other. The derived structural parameters are in very good agreement in all bands. 115 02 09 62 17 20 61 63 10 35 60 16 25 01 26 59 01 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 12 14 91 36 37 08 06 00 53 98 02 46 85 96 22 84 ...... 02 14 08 21 64 48 12 17 14 17 25 17 30 17 08 14 25 21 65 48 14 5 17 5 01 0 69 6 44 5 01 0 ...... GALFIT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 25 81 44 55 60 89 89 17 11 22 52 50 86 41 65 85 ...... 01 15 04 21 99 48 41 18 39 18 09 17 10 17 05 15 16 22 79 48 07 4 09 4 01 0 50 5 26 4 01 0 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 46 10 33 06 07 54 51 32 60 24 32 68 89 20 85 86 ...... 01 15 03 22 30 48 13 19 16 19 13 17 15 17 05 15 15 22 01 48 01 4 08 4 01 0 41 6 24 4 01 0 ...... 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± BVRI 39 07 38 95 00 05 04 33 27 68 00 75 91 43 02 88 ...... 22 16 02 23 06 48 07 18 06 18 10 16 62 23 76 4 14 0 28 5 31 4 08 0 ...... 37 48 0 1 0 0 0 0 0 . 3 0 2 1 0 6 ± ± ± ± ± ± ± ± ± ± ± ± ± NUV 4.00* 4 82 47 73 39 35 65 82 63 51 03 57 50 ...... 0 3 0 model. 4 / 1 ) 24 ) 24 r 2 2 − − Table 5.2: Structural parameters of PGC 1000714 obtained by (mag) ... 19 (mag) ... 20 (mag) 17 (mag) 17 (mag) 17 (mag) 17 E C E C (arcsec) 6 (arcsec) 8 (mag arcsec (mag arcsec total total total total total total e e e e CORE CORE INNER RING OUTER RING § m µ R n b/a P.A. (degree) 23 § m m § m m † m µ R n b/a P.A. (degree) 25 The initial parameters areThe fixed initial to parameters the are 1D allowed fitting to results. vary. § † * The core is fitted by the 116 35 34 20 26 . . . . 0 0 0 0 ± ± ± ± 30 32 46 50 . . . . 28 18 30 18 17 17 22 17 . . . . 0 0 0 0 ± ± ± ± 32 38 00 01 . . . . GALFIT (continued) 20 19 28 19 03 18 06 18 . . . . 0 0 0 0 ± ± ± ± 03 22 91 91 . . . . 22 20 25 20 06 17 08 17 . . . . 0 0 0 0 ± ± ± ± BVRI 85 85 14 12 . . . . 03 18 03 18 . . 0 0 ± ± NUV 41 40 . . (mag) ... 20 (mag) ... 20 (mag) 17 (mag) 17 E C E C Table 5.2: Structural parameters of PGC 1000714 obtained by total total total total INNER RING OUTER RING † m m † m m 117 5.6 Rings

The outer ring can be seen well in the NUV and BVRI bands (see Figure 5.2). While NIR imaging is not optimal for studying the outermost part of the galaxy, the outer ring cannot be detected in the MIR data. In Section 5.4.1, the aim was to study the geometric properties of the core (e.g., the ǫ and P.A. profiles), so ǫ and P.A. were allowed to vary. Due to the clumpy and non-uniform structure of the ring, the unconstrained ellipse fitting procedure cannot be applied on the outer ring. To study the structure of the outer ring, the ellipse fitting is repeated, but this time the program is forced to extend the fit well beyond the core with the fixed ǫ and P.A. of the outer ring, which is the ellipticity (ǫ = 0.2) and orientation (P.A.= 60◦) of the white ellipses in Figure

−2 5.4. It is found that the mean surface brightness drops to µB = 26.8 mag arcsec at the faintest level of the core, at a radius of 21.5 arcsec. The outer ring has no sharp ∼ outer boundary and fades away at a radius of 40 arcsec. The peak surface brightness ∼ −2 of the outer ring is µB = 25.9 mag arcsec at r = 30 arcsec and it is fainter than that of Hoag’s object (Finkelman et al. (2011) reported the peak surface brightness of

−2 the outer ring for Hoags object is µB = 24.6 mag arcsec at r = 19.1 arcsec). In the case of the NUV band, the core profile was fitted by an r1/4 model. The S´ersic bulge model was not fitted to the NUV data due to the large resolution. The light profile of the galaxy with the outer ring is shown in Figure 5.10, where the lower panel shows the residuals after the best-fitting core model in Figure 5.7 is subtracted. The residuals reveal an inner ring at radius between 8 arcsec and 15 arcsec, which is very faint ∼ ∼ in the blue light, and an outer ring at radius between 20 and 40 arcsec, which is very bright in the NUV light. Unfortunately, the nature of the faint inner ring cannot be verified in the MIR data due to the large PSF effects. To estimate the total intensity of the rings, the residual images generated by GAL- FIT (see Figure 5.9) were used. Using the residual images, total intensities of the outer 118

Figure 5.10: The light profiles of the galaxy with the outer ring are shown for the NUV and BVRI bands. The lower panels show the residuals after the best-fitting core model in Figure 5.7 is subtracted. The residuals reveal an inner ring at a radius between 8 ∼ and 15 arcsec: It is very faint in the blue light and gets brigher towards redder light. ∼ In addition, the residuals show an outer ring at a radius between 20 and 40 arcsec: ∼ ∼ It is very bright in the NUV light and gets fainter towards redder light. 119 ring were derived from the region between 20 and 40 arcsec while the total intensity from the region inside 20 arcsec was adopted as that of the inner ring. The region was enclosed by the same white ellipses in Figure 5.4. The total intensities of the rings were also derived from the region defined by circles. These intensities were then con- verted into magnitudes, using proper zero-point magnitudes. The derived magnitudes are listed in Table 5.2.

5.7 Stellar Populations

SED fitting is a common technique for deriving galaxy properties, and it relies on comparing the observed SED to a set of model SEDs and searching for the best match. Since the physical properties of the models are known, this knowledge can be used to recover the properties of an observed galaxy. In order to study the stellar populations of the core and the outer ring, the HYPERZ SED fitting code, version 1.2 (Bolzonella et al. 2000) was used. Due to the large uncertainty on the nature of the inner ring in the infrared light, an attempt was not made to constraint its age. The observed SED for the bulge and the outer ring was defined by the mean of the magnitudes listed in Table 2, and the standard deviation of the values was used to estimate the uncertainty (see Table 5.3). The SED fitting method was based on minimization of χ2, which was computed for templates covering a broad range of star formation modes, ages and absorption values. The default was an age grid of 51 values (minimum and maximum age are 32 Myr and 19.5 Gyr, respectively). Galaxy age was constrained to be younger than the age of the Universe. The best-fitting solution yielded the stellar age, star formation law, and dust reddening. The set-up consisted of the mean spectra of local galaxies from Coleman et al. (1980), and Bruzual A. and Charlot (1993) evolving synthesis models. 120

Table 5.3: The observed SED of the galaxy components in units of magnitude.

Filters Bulge Outer ring Near-UV NUV 17.74 0.24 17.39 0.11 ± ± B 16.36 0.06 18.09 0.23 ± ± V 15.39 0.11 17.72 0.27 Optical ± ± R 15.21 0.10 17.95 0.48 ± ± I 14.06 0.13 17.28 0.97 ± ± J 13.34 0.52 ... ± Near-IR H 12.80 0.44 ... ± Ks 12.51 0.34 ... ± W1 12.56 0.35 ... ± W2 12.61 0.30 ... Mid-IR* ± W3 11.64 0.44 ... ± W4 ......

* The photometric information from the WISE Source Catalog is used.

Three different star formation models are considered: single burst, constant star for- mation and exponentially decaying star formation histories (SFH) where star formation rate (SFR) is defined as SFR(t) = (1/τ) exp( t/τ) in which the τ parameter is the − “time-scale” when the star formation was most intense with τ = 1, 2, 3, 5, 15, 30 Gyr, as included in the HYPERZ package. The attenuation law from Calzetti et al. (2000) was used. Solar metallicity and a Salpeter (1955) initial mass function (IMF) with lower and upper mass cut-offs of 0.1M⊙ and 100M⊙ were assumed. In addition to this set-up, I also used a self-consistent set of templates, where the evolution in metallicity of stellar population is explicitly taken into account (see Mobasher and Mazzei 1999), as included in the HYPERZ package. The best-fitting solution for the bulge is obtained with an exponentially decaying SFH with τ = 1 Gyr and solar metallicity. The age of the best template for the bulge is 5.5 Gyr (falling between 4.5 Gyr and 6.5 Gyr in the age grid). The observed SED of the outer ring is well reproduced by a stellar population with 121 solar metallicity and an instantaneous burst scenario. The age of the best template for the ring is 0.1278 Gyr (falling between 0.09048 Gyr and 0.18053 Gyr in the age grid). Figure 5.11 shows the observed SED of the bulge and the ring with their best-fitting models. Identical results are obtained when different IMFs (namely Miller and Scalo 1979 and Chabrier 2003) are used in the set-up.

Figure 5.11: The observed SED of the core and the ring (red circles) with their best- fitting models (black solid lines) are shown. The y-axes show the flux units in AB magnitudes.

5.8 Discussion

It has been shown that PGC 1000714 presents an enhancement of the blue light in a fairly circular region centered on the core. The detailed photometry reveals an elliptical galaxy-like core with a diffuse reddish inner ring-shaped structure, but no sign of a bar or stellar disk. The SED fitting suggests that the galaxy is fairly young: the core being ∼ 5.5 Gyr old, and the detached outer ring being 0.13 Gyr old. The central body is well ∼ matched with an exponential decaying SFR with τ = 1 Gyr, that produces a reasonable fit of the photometric properties of elliptical galaxies in the local Universe. A shallow intrinsic metallicity gradient along the core (see Figure 5.6) is consistent with a major 122 merger scenario, in which a merging event causes the turbulent mixing and flattens any possible ambient gradient present in the progenitor galaxies (White 1980). It is unlikely that the core is the result of a monolithic gas collapse. According to this model, a rapid dissipative collapse forms stars while the gas sinks to the center of the forming galaxy (Larson 1974, 1975; Carlberg 1984; Arimoto and Yoshii 1987). The infalling gas, which is chemically enriched by evolving stars, contributes to the metal-rich star formation in the center, thus establishing very steep color gradients (S´anchez-Bl´azquez et al. 2006; Spolaor et al. 2009, and references therein). The shallow color gradient along the core contradicts the predictions of the monolithic collapse model. The SED fitting deduces clearly different ages, stellar characteristics and SFHs for the outer ring and the core. Based on the color of the rings, different formation histories are anticipated for the inner and outer ring, however one cannot determine the formation mechanism for the diffuse inner ring with confidence based on the data presented in this work. To say more about the formation of the inner ring, infrared data of PGC 1000714 with better resolution is needed. A number of ideas may be put forward to explain the outer ring in such a pe- culiar galaxy, but the observational results rule out most of these. In the widely accepted bar-driven ring model, one may expect that PGC 1000714 was a barred early-type galaxy once, where it experienced a strong bar instability in the disk, an outer ring formed in the process, and the bar and disk were subsequently de- stroyed. A vertical buckling instability in the disk may indeed weaken the bar and dissolve the outer half of the bar (Martinez-Valpuesta et al. 2006) within a Gyr or less (Athanassoula and Mart´ınez-Valpuesta 2008). However, the buckling instability is not expected to lead to a complete bar dissolution (Martinez-Valpuesta and Shlosman 2004). As reviewed by Buta and Combes (1996), if a satellite merger is the driving force for the bar destruction, the effective dissolution of the bar may occur in about 20 Myr 123 (Pfenniger 1991). However, such a process is expected to cause some non-axisymmetric features on the central body, which has not been detected in PGC 1000714. If gas infall towards the center is the reason for the bar destruction, the decoupling of a secondary bar might occur in the dissolution of the primary one when there is a high accretion rate (Friedli and Martinet 1993). For a milder gas accretion rate, the time-scale for the bar dissolution can be very long due to self-regulation (Buta and Combes 1996). In addition, the bar destruction process affects the orbital structures and may cause lens formation (Combes 1996), which is not detected in PGC 1000714. Moreover, such a process is also expected to create a light profile of the core with a low S´ersic index of n < 2 because pseudobulges are expected to retain fingerprints of their disky ori- gin (Kormendy and Kennicutt 2004). Yet the core has no sign of a disk, and presents properties of a spheroid with a smooth r1/4 light distribution for 6.0 mag. Note also that the rings of barred galaxies are much more luminous than that of PGC 1000714 (M = 16.9 0.1), e.g., M 20.5, 20.3, and 19.6 for NGC 1291, NGC 2217, B − ± B ∼ − − − and NGC 2859 (Sandage and Tammann 1981). Debattista et al. (2004) used collionless N-body simulations to study the final properties of disks that suffer a bar instability, and failed to reproduce round pseudobulges inside low-inclination galaxies, as in PGC 1000714. As pointed out by Schweizer et al. (1987) for Hoag’s Object, a disk is needed to give rise to a bar instability. However, it is worth mentioning that Gadotti and de Souza (2003) proposed a mechanism to form bars in spheroids through the dynamical effects of a sufficiently eccentric halo, without the need of a stellar disk. To say more about this model, the information about the halo structure of the galaxy is needed. Galaxy interactions can be considered as an alternative mechanism for the outer ring formation. In the galaxy-galaxy collision model, PGC 1000714 can be considered as a classical ring galaxy with its companion superimposed. High-resolution numerical simulations of binary mergers of disk and early-type galaxies find most of the fast 124 rotators formed due to major mergers have intermediate ellipticities between 0.4 and 0.6, and mostly contain a bar (Bois et al. 2011). However, it is found that PGC 1000714 has a fairly round central body with a fairly round inner ring and no trace of a bar. Moreover, the geometrical properties derived from the optical images stay constant along the central core (e.g., ǫ 0.1 and P.A. 50◦). The smooth profile in the outer ≈ ≈ region implies that the face-on ring is in equilibrium with the host galaxy and that the system has settled to its current configuration. Since the initial conditions must be fine-tuned in order to create a polar ring from a merger event (Macci`oet al. 2006), it is highly unlikely this will create the peculiar properties of PGC 1000714 from a merger event. Therefore, the possibility of the galaxy having a history of a major collision is ruled out. A close passage of a nearby galaxy or a tidally interacting companion can mimic the effect of a bar by creating spiral structure or a ring in non-barred galaxies (Elmegreen and Elmegreen 1983; Combes 1988; Tutukov and Fedorova 2006). This non-axisymmetric presence can trigger a recent burst of star formation and create a UV-bright ring structure in the outer disks of spirals (Thilker et al. 2007). Although an orbiting small companion could create a detached outer ring, such an external perturba- tion could also destroy an existing ring, break a true ring into a pseudo-ring structure, or prevent it from occuring in the first place (Buta and Combes 1996). Together with the failure in finding a possible companion as well as detecting any tidal tail, shells or ripple signature brighter than µ 27 mag arcsec−2, it is unlikely that the tidal action B ≈ is the origin of the outer ring in PGC 1000714. An accretion model can also explain the formation of a detached outer ring. This model is used to explain the formation of narrow PRGs, which share similar charac- teristics with PGC 1000714. As an example, the core of narrow PRG AM 2020-504 follows an r1/4 law well over a large range of surface brightnesses (Arnaboldi et al. 125 1993; Iodice et al. 2002) and its polar ring appears as a peak on the underlying central body (Iodice 2014). The age of the central body and the ring of AM 2020-504 (3-5 Gyr and 1 Gyr, respectively; Iodice 2014) are comparable to those of PGC 1000714. As a formation mechanism for AM 2020-504, Iodice (2014) suggested tidal accretion of material from outside by a pre-existing early-type galaxy since this process would not modify the global structure of the progenitor accreting galaxy, whose morphology remains spheroidal with the same colors and age. Due to the similar characteristics, this argument is also valid for PGC 1000714. However, it is worth mentioning that the radius of the outer ring in PGC 1000714 is not small with respect to the semi-major axis radius of the central body as in the narrow PRGs, hence PGC 1000714 cannot be a narrow PRG with the polar structure moderately inclined to the line of sight (nearly face-on). The exponential-like light profile of the polar structure in the wide PRGs (Reshetnikov et al. 1994; Iodice et al. 2002; Spavone et al. 2012) is also not consistent with that in PGC 1000714. On the other hand, the absolute magnitude of the outer ring (M = 16.9 0.1) is comparable to those of PRGs, e.g., M = 18.2 0.5, B − ± B − ± 17.6 0.6, and 15.5 1.0 for the wide PRG NGC 4650A, the wide PRG A0136-0801, − ± − ± and the narrow PRG ESO 415-G26, respectively (Whitmore et al. 1987). Peculiar galax- ies with several decoupled ring-like structures are also observed, in which at least one of them is in the polar direction with respect to the central host, e.g., ESO 474-G26 (Reshetnikov et al. 2005; Spavone et al. 2012). While the central body of ESO 474- G26 is an elliptical-like object (and almost round; ǫ 0.06), it is much brighter than ∼ that of PGC 1000714. Also, both rings in ESO 474-G26 have similar colors, and they are very irregular (Spavone et al. 2012). The simulations suggest that the structure of ESO 474-G26 could be a transient phase and not a stable dynamical configuration (Reshetnikov et al. 2005; Spavone et al. 2012). Spavone et al. (2012) ruled out the grad- ual disruption of a dwarf satellite galaxy as a formation mechanism for this multiple ring 126 structure. However, note that the rings of PGC 1000714, which have different colors, require different formation mechanisms, therefore the argument made by Spavone et al. (2012) is not valid for it. Also, unlike ESO 474-G26, the smooth geometric profile at the outer region of the core implies that the face-on ring is in equilibrium with the host galaxy and that the system has settled into its current configuration. No kinematic data is available for PGC 1000714, therefore one cannot verify whether it is a PRG or not. However, it shares several characteristics with them, and therefore, may be formed by a similar formation mechanism. In addition, the accretion model is also used to explain the formation of the outer rings in Hoag-type galaxies (Schweizer et al. 1987; Finkelman and Brosch 2011). In Hoag’s Object, the complete regularity of the central body is suggested to be a result of a recent accretion event (Schweizer et al. 1987). Al- though PGC 1000714 is not a true Hoag-type galaxy due to the inner diffuse ring, the outer ring may be formed by a similar mechanism, based on the characteristic similar- ities. Table 5.4 summarizes the comparison between PGC 1000714, the narrow PRG AM 2020-504 (Iodice et al. 2002), the double ringed PRG ESO 474-G26 (Spavone et al. 2012), Hoag’s Object (Finkelman and Brosch 2011; Finkelman et al. 2011), and UGC 4599 (a Hoag-type galaxy defined by Finkelman and Brosch 2011). Given that the colors of the outer ring imply a 0.13 Gyr stellar population, regularity of the central body of PGC 1000714 with no tidal signature implies that the galaxy could have undergone a recent accretion event as proposed in PRGs and Hoag-type galaxies. It is possible the gas could have been accreted from a single event such as a passing gas- rich dwarf companion. Previous studies of gas-rich, early-type galaxies have shown that nearby companions are often missed by optical catalogs or show no sign of an optical counterpart (e.g., Serra et al. 2006; Struve et al. 2010). A study on the HI environment of PGC 1000714 is therefore needed to complement this study. 127 02 04 02 02 . 04 04 02 05 04 04 14 ...... 0 0 0 0 9 0 0 0 0 0 0 0 . 1 5 ± 0 ± ± ± ± ± ± ± ± ± ± 9 ∼ > . ∼ 98 26 18 17 38 68 97 16 34 81 ...... 17 − − 02 02 05 11 02 14 07 15 . . 04 0 04 0 07 1 10 0 12 0 14 0 ...... 0 0 0 0 0 0 0 0 0 0 0 1 2 10 ± ± ± ± ± ± ± ± ± ± ± 8 1 ∼ < . . > 58 42 12 42 86 44 51 26 96 ...... 19 20 − − 97 ...... 67 ...... 06 ...... 02 13 . . 08 ...... 04 0 . . . . 0 0 9 7.8 2.4 2 ...... 16 0 16 0 . . 1 29 15.5 7.7 19 ...... 1 1 ± ± − ± − ± ∼ ∼ ∼ ∼ ∼ 94 51 50 55 30 12 ...... 0-504 (Iodice et al. 2002; Iodice 2014), the double 11). al. 2005), Hoag’s Object (Finkelman and Brosch 2011; 40 15 . 18 75 3 . 0 − ∼ ∼ 96 . 02 ...... 02 ...... 06 16.63 13 09 17 52 13.50 12 33 ... 15 11 ...... 16 27 ...... 16 21 ...... 2 04 ...... 0 05 1.04 1 05 ...... 0 0 3-5 1 ...... 0 0 05 3.98 2 05 ...... 0 04 ...... 0 . . . . . 0 0 0 0 0 0 . . . 5 ... 0 0 0 0 1 85 . 0 0 0 6 ... ± ± . 2 ± ± ± ± ± ± 0 ± ± ± ± ± 6.5 0.75 ± ± ± ∼ 43 07 5 ∼ . . 36 13 2 1 34 2 85 39 72 . ∼ 42 21 95 59 ...... 4 1 0 19 17 2 0 0 0 16 18 13 20 15 17 PGC 1000714 AM 2020-504 ESO 474-G26 Hoag’s Object UGC 4599 − − ] (kpc) 15.6 ] (kpc) 1 2 V e e − K R R I I K R R (kpc) 2.4 4.0* 15.3 2.0 3.2 / / V V − − − − − B B J B V V − e 1 2 R R m m b/a B R Radius [R Radius [R R m V M m M NUV m m V J R Age (Gyr) 0.13 1 0.1-0.03 Age (Gyr) 5 Bulge Ring 1 Ring 2 * Arnaboldi et al. (1993) ringed-PRG ESO 474-G26 (Spavone etFinkelman et al. 2012; al. Reshetnikov 2011), et and UGC 4599 (Finkelman and Brosch 20 Table 5.4: Comparison between PGC 1000714, narrow PRG AM 202 128 5.9 Conclusions

This chapter presents a photometric study of PGC 1000714, a galaxy with a superfi- cial resemblance to Hoag’s Object. The detailed 2D image analysis reveals that PGC 1000714 is a genuine non-barred galaxy in a low-density environment with two rings. Even with the NIR bands, I have failed to detect any trace of a bar-like structure. The geometric properties derived from the optical images, in particular ǫ and P.A., stay constant along the central core. The smooth profiles at the outer region imply that the system has settled into its current configuration. The nearly round central body follows closely a r1/4 light profile for more than 5.0 mag almost all the way to the center. Based on the 2D B I color index map and the photometric decomposition presented − in this work, the galaxy is classified as a double ringed elliptical (E2) galaxy, which was misclassified as (R)SAa in the RC3. Without the kinematic data, one cannot confirm or reject the possibility of PGC 1000714 being a PRG. If the galaxy is really a PRG, it is a very peculiar example of this unique class of objects. If it is not a PRG, this current work has revealed, for the first time, that an elliptical galaxy has two fairly round rings. Studying such peculiar galaxies is important to address how different kinds of interac- tions (i.e., internal or external such as galaxy-galaxy and galaxy-environment) lend to different galaxy morphologies. A number of formation scenarios are discussed based on the photometric results. A careful inspection of the color profiles reveals a fairly constant color index along the core. The inner ring is very faint in blue light and gets brighter towards redder light while the outer ring is very bright in UV light and gets fainter towards redder light. The SED fitting suggests the galaxy is fairly young such that the core is 5.5 Gyr old while the outer ring is just 0.13 Gyr old. The outer ring is well fitted by a stellar population with solar metallicity and an instantaneous burst scenario. However, the best-fitting solution for the central body is obtained by an exponentially decaying SFH and solar 129 metallicity. This is consistent with the photometric properties of elliptical galaxies in the local Universe. Based on the color of the rings, different formation histories are anticipated for the inner and outer ring. To recover the formation history of the inner ring, infrared data of PGC 1000714 with better resolution is needed. Among several ways to create the outer ring in a non-barred galaxy, it is concluded that a recent accretion event such as accretion from a gas-rich is the most plausible scenario that accounts for the observed properties of the galaxy. Spectroscopic data and a study of the HI environment of PGC 1000714 are needed to say more about the evolution of this galaxy in a low-density environment. References

F. C. Adams, D. S. Graff, and D. O. Richstone. A Theoretical Model for the MBH -σ Relation for Supermassive Black Holes in Galaxies. ApJl, 551:L31–L35, April 2001.

P. D. Allen, S. P. Driver, A. W. Graham, E. Cameron, J. Liske, and R. de Propris. The Millennium Galaxy Catalogue: bulge-disc decomposition of 10095 nearby galaxies. MNRAS, 371:2–18, September 2006.

P. N. Appleton and C. Struck-Marcell. Collisional Ring Galaxies. Fund. Cosmic Physics, 16:111–220, 1996.

N. Arimoto and Y. Yoshii. Chemical and photometric properties of a galactic wind model for elliptical galaxies. A&A, 173:23–38, February 1987.

M. Arnaboldi, M. Capaccioli, E. Cappellaro, E. V. Held, and L. Sparke. Studies of narrow polar rings around E galaxies. I - Observations and model of AM 2020-504. A&A, 267:21–30, January 1993.

M. Arnaboldi, T. Oosterloo, F. Combes, K. C. Freeman, and B. Koribalski. New H I Observations of the Prototype Polar Ring Galaxy NGC 4650A. AJ, 113:585–598, February 1997.

E. Athanassoula. Evolution of Bars in Isolated and in Interacting Disk Galaxies. In R. Buta, D. A. Crocker, and B. G. Elmegreen, editors, IAU Colloq. 157: Barred

130 131 Galaxies, volume 91 of Astronomical Society of the Pacific Conference Series, page 309, 1996.

E. Athanassoula and A. Bosma. Shells and rings around galaxies. ARA&A, 23:147–168, 1985.

E. Athanassoula and I. Mart´ınez-Valpuesta. Formation and Evolution of Bulges. In J. H. Knapen, T. J. Mahoney, and A. Vazdekis, editors, Pathways Through an Eclectic Universe, volume 390 of Astronomical Society of the Pacific Conference Series, page 454, June 2008.

E. Athanassoula, S. Morin, H. Wozniak, D. Puy, M. J. Pierce, J. Lombard, and A. Bosma. The shape of bars in early-type barred galaxies. MNRAS, 245:130–139, July 1990.

M. Bachetti, V. Rana, D. J. Walton, D. Barret, F. A. Harrison, S. E. Boggs, F. E. Christensen, W. W. Craig, A. C. Fabian, F. F¨urst, B. W. Grefenstette, C. J. Hailey, A. Hornschemeier, K. K. Madsen, J. M. Miller, A. F. Ptak, D. Stern, N. A. Webb, and W. W. Zhang. The Ultraluminous X-Ray Sources NGC 1313 X-1 and X-2: A Broadband Study with NuSTAR and XMM-Newton. ApJ, 778:163, December 2013.

M. Bachetti, F. A. Harrison, D. J. Walton, B. W. Grefenstette, D. Chakrabarty, F. F¨urst, D. Barret, A. Beloborodov, S. E. Boggs, F. E. Christensen, W. W. Craig, A. C. Fabian, C. J. Hailey, A. Hornschemeier, V. Kaspi, S. R. Kulkarni, T. Mac- carone, J. M. Miller, V. Rana, D. Stern, S. P. Tendulkar, J. Tomsick, N. A. Webb, and W. W. Zhang. An ultraluminous X-ray source powered by an accreting neutron star. Nature, 514:202–204, October 2014.

M. Baes, P. Buyle, G. K. T. Hau, and H. Dejonghe. Observational evidence for a 132 connection between supermassive black holes and dark matter haloes. MNRAS, 341: L44–L48, June 2003.

V. F. Baldassare, A. E. Reines, E. Gallo, and J. E. Greene. A ˜50, 000M Black Hole in the Nucleus of RGG 118. ApJl, 809:L14, August 2015.

K. Bandara, D. Crampton, and L. Simard. A Relationship Between Supermassive Black Hole Mass and the Total Gravitational Mass of the Host Galaxy. ApJ, 704:1135–1145, October 2009.

A. J. Barth. Black hole masses in active galaxies. In T. Storchi-Bergmann, L. C. Ho, and H. R. Schmitt, editors, The Interplay Among Black Holes, Stars and ISM in Galactic Nuclei, volume 222 of IAU Symposium, pages 3–8, November 2004.

A. J. Barth, L. C. Ho, R. E. Rutledge, and W. L. W. Sargent. POX 52: A Dwarf Seyfert 1 Galaxy with an Intermediate-Mass Black Hole. ApJ, 607:90–102, May 2004.

A. J. Barth, J. E. Greene, and L. C. Ho. Low-Mass Seyfert 2 Galaxies in the Sloan Digital Sky Survey. AJ, 136:1179–1200, September 2008.

H. Baumgardt, P. Hut, J. Makino, S. McMillan, and S. Portegies Zwart. On the Central Structure of M15. ApJl, 582:L21–L24, January 2003.

A. Beifiori, S. Courteau, E. M. Corsini, and Y. Zhu. On the correlations between galaxy properties and supermassive black hole mass. MNRAS, 419:2497–2528, January 2012.

K. Bekki. Formation of Polar-Ring S0 Galaxies in Dissipative Galaxy Mergers. ApJl, 490:L37–L40, November 1997.

K. Bekki. Unequal-Mass Galaxy Mergers and the Creation of Cluster S0 Galaxies. ApJl, 502:L133–L137, August 1998. 133 M. Bernardi, R. K. Sheth, J. Annis, S. Burles, D. J. Eisenstein, D. P. Finkbeiner, D. W. Hogg, R. H. Lupton, D. J. Schlegel, M. SubbaRao, N. A. Bahcall, J. P. Blakeslee, J. Brinkmann, F. J. Castander, A. J. Connolly, I. Csabai, M. Doi, M. Fukugita, J. Frieman, T. Heckman, G. S. Hennessy, Z.ˇ Ivezi´c, G. R. Knapp, D. Q. Lamb, T. McKay, J. A. Munn, R. Nichol, S. Okamura, D. P. Schneider, A. R. Thakar, and D. G. York. Early-type Galaxies in the Sloan Digital Sky Survey. II. Correlations between Observables. AJ, 125:1849–1865, April 2003.

M. Bernardi, A. Meert, R. K. Sheth, V. Vikram, M. Huertas-Company, S. Mei, and F. Shankar. The massive end of the luminosity and stellar mass functions: dependence on the fit to the light profile. MNRAS, 436:697–704, November 2013.

J. C. Berrier, B. L. Davis, D. Kennefick, J. D. Kennefick, M. S. Seigar, R. S. Barrows, M. Hartley, D. Shields, M. C. Bentz, and C. H. S. Lacy. Further Evidence for a Supermassive Black Hole Mass-Pitch Angle Relation. ApJ, 769:132, June 2013.

E. Bertin and S. Arnouts. SExtractor: Software for source extraction. A&AS, 117: 393–404, June 1996.

M. R. Blanton, D. W. Hogg, N. A. Bahcall, J. Brinkmann, M. Britton, A. J. Connolly, I. Csabai, M. Fukugita, J. Loveday, A. Meiksin, J. A. Munn, R. C. Nichol, S. Okamura, T. Quinn, D. P. Schneider, K. Shimasaku, M. A. Strauss, M. Tegmark, M. S. Vogeley, and D. H. Weinberg. The Galaxy Luminosity Function and Luminosity Density at Redshift z = 0.1. ApJ, 592:819–838, August 2003.

P. Boggs and J. Rogers. Orthogonal Distance Regression. Contemporary Mathematics, 112:183–194, June 1990.

M. Bois, E. Emsellem, F. Bournaud, K. Alatalo, L. Blitz, et al. The ATLAS3D project - 134 VI. Simulations of binary galaxy mergers and the link with fast rotators, slow rotators and kinematically distinct cores. MNRAS, 416:1654–1679, September 2011.

M. Bolzonella, J.-M. Miralles, and R. Pell´o. Photometric redshifts based on standard SED fitting procedures. A&A, 363:476–492, November 2000.

C. M. Booth and J. Schaye. Cosmological simulations of the growth of supermassive black holes and feedback from active galactic nuclei: method and tests. MNRAS, 398: 53–74, September 2009.

C. M. Booth and J. Schaye. Dark matter haloes determine the masses of supermassive black holes. MNRAS, 405:L1–L5, June 2010.

F. Bournaud and F. Combes. Formation of polar ring galaxies. A&A, 401:817–833, April 2003.

F. Bournaud, C. J. Jog, and F. Combes. Galaxy mergers with various mass ratios: Properties of remnants. A&A, 437:69–85, July 2005.

R. G. Bower, A. J. Benson, R. Malbon, J. C. Helly, C. S. Frenk, C. M. Baugh, S. Cole, and C. G. Lacey. Breaking the hierarchy of galaxy formation. MNRAS, 370:645–655, August 2006.

C. B. Brook, F. Governato, T. Quinn, J. Wadsley, A. M. Brooks, B. Willman, A. Stilp, and P. Jonsson. The Formation of Polar Disk Galaxies. ApJ, 689:678–686, December 2008.

N. Brosch. The nature of Hoag’s object - The perfect ringed galaxy. A&A, 153:199–206, December 1985.

G. Bruzual A. and S. Charlot. Spectral evolution of stellar populations using isochrone synthesis. ApJ, 405:538–553, March 1993. 135 J. S. Bullock, T. S. Kolatt, Y. Sigad, R. S. Somerville, A. V. Kravtsov, A. A. Klypin, J. R. Primack, and A. Dekel. Profiles of dark haloes: evolution, scatter and environ- ment. MNRAS, 321:559–575, March 2001.

M. Bureau and E. Athanassoula. Bar Diagnostics in Edge-On Spiral Galaxies. III. N-Body Simulations of Disks. ApJ, 626:159–173, June 2005.

M. Bureau and K. C. Freeman. The Nature of Boxy/Peanut-Shaped Bulges in Spiral Galaxies. AJ, 118:126–138, July 1999.

R. Buta. The Structure and Dynamics of Ringed Galaxies. II. UBVRI Surface Pho- tometry and H alpha Kinematics of the Ringed Barred Spiral NGC 1433. ApJS, 61: 631–665, August 1986.

R. Buta. Weakly barred early-type ringed galaxies. IV - The double-ringed SO(+) galaxy NGC 7702. ApJ, 370:130–139, March 1991.

R. Buta and F. Combes. Galactic Rings. Fund. Cosmic Phys., 17:95–281, 1996.

D. Calzetti, L. Armus, R. C. Bohlin, A. L. Kinney, J. Koornneef, and T. Storchi- Bergmann. The Dust Content and Opacity of Actively Star-forming Galaxies. ApJ, 533:682–695, April 2000.

R. G. Carlberg. Dissipative models for the sequence of elliptical galaxies. ApJ, 286: 416–421, November 1984.

V. Casasola, F. Combes, S. Garc´ıa-Burillo, L. K. Hunt, S. L´eon, and A. J. Baker. Molecular gas in NUclei of GAlaxies (NUGA). X. The Seyfert 2 galaxy NGC 3147. A&A, 490:61–76, October 2008.

A. Cattaneo. Quasars and galaxy formation. MNRAS, 324:128–140, June 2001. 136 A. Cattaneo, M. G. Haehnelt, and M. J. Rees. The distribution of supermassive black holes in the nuclei of nearby galaxies. MNRAS, 308:77–81, September 1999.

G. Chabrier. Galactic Stellar and Substellar Initial Mass Function. PASP, 115:763–795, July 2003.

K. T. E. Chua, A. Pillepich, V. Rodriguez-Gomez, M. Vogelsberger, S. Bird, and L. Hernquist. Subhalo Demographics in the Illustris Simulation: Effects of Baryons and Halo-to-Halo Variation. ArXiv e-prints, November 2016.

B. C. Ciambur. Profiler - A Fast and Versatile New Program for Decomposing Galaxy Light Profiles. PASA, 33:e062, December 2016.

L. Ciotti, J. P. Ostriker, and D. Proga. Feedback from Central Black Holes in Elliptical Galaxies. I. Models with Either Radiative or Mechanical Feedback but not Both. ApJ, 699:89–104, July 2009.

G. D. Coleman, C.-C. Wu, and D. W. Weedman. Colors and magnitudes predicted for high redshift galaxies. ApJS, 43:393–416, July 1980.

F. Combes. Influence of Large-Scale Dynamics Upon Star Formation in Galaxies. In R. E. Pudritz and M. Fich, editors, NATO Advanced Science Institutes (ASI) Series C, volume 232 of NATO Advanced Science Institutes (ASI) Series C, page 475, 1988.

F. Combes. Ring and Lens Formation. In R. Buta, D. A. Crocker, and B. G. Elmegreen, editors, IAU Colloq. 157: Barred Galaxies, volume 91 of Astronomical Society of the Pacific Conference Series, page 286, 1996.

F. Combes. Polar Ring Galaxies and Warps. In G. A. Mamon, F. Combes, C. Deffayet, and B. Fort, editors, EAS Publications Series, volume 20 of EAS Publications Series, pages 97–104, 2006. 137 F. Combes and M. Arnaboldi. The dark halo of polar-ring galaxy NGC 4650a: flattened towards the polar ring? A&A, 305:763, January 1996.

F. Combes and M. Gerin. Spiral structure of molecular clouds in response to bar forcing - A particle simulation. A&A, 150:327–338, September 1985.

C. J. Conselice, M. A. Bershady, M. Dickinson, and C. Papovich. A Direct Measurement of Major Galaxy Mergers at z < 3. AJ, 126:1183–1207, September 2003.

D. J. Croton, V. Springel, S. D. M. White, G. De Lucia, C. S. Frenk, L. Gao, A. Jenkins, G. Kauffmann, J. F. Navarro, and N. Yoshida. The many lives of active galactic nuclei: cooling flows, black holes and the luminosities and colours of galaxies. MNRAS, 365: 11–28, January 2006.

B. L. Davis, J. C. Berrier, D. W. Shields, J. Kennefick, D. Kennefick, M. S. Seigar, C. H. S. Lacy, and I. Puerari. Measurement of Galactic Logarithmic Spiral Arm Pitch Angle Using Two-dimensional Fast Fourier Transform Decomposition. ApJS, 199:33, April 2012.

B. L. Davis, J. C. Berrier, L. Johns, D. W. Shields, M. T. Hartley, D. Kennefick, J. Kennefick, M. S. Seigar, and C. H. S. Lacy. The Black Hole Mass Function Derived from Local Spiral Galaxies. ApJ, 789:124, July 2014.

B. L. Davis, A. W. Graham, and M. S. Seigar. ApJ, submitted, 2017.

R. S. de Jong. Near-infrared and optical broadband surface photometry of 86 face-on disk dominated galaxies. II. A two-dimensional method to determine bulge and disk parameters. A&AS, 118:557–573, September 1996.

R. E. de Souza, D. A. Gadotti, and S. dos Anjos. BUDDA: A New Two-dimensional 138 Bulge/Disk Decomposition Code for Detailed Structural Analysis of Galaxies. ApJS, 153:411–427, August 2004.

G. de Vaucouleurs, A. de Vaucouleurs, H. G. Corwin, Jr., R. J. Buta, G. Paturel, and P. Fouqu´e. Third Reference Catalogue of Bright Galaxies. Volume I: Explanations and references. Volume II: Data for galaxies between 0h and 12h. Volume III: Data for galaxies between 12h and 24h. 1991.

P. T. de Zeeuw, M. Bureau, E. Emsellem, R. Bacon, C. M. Carollo, Y. Copin, R. L. Davies, H. Kuntschner, B. W. Miller, G. Monnet, R. F. Peletier, and E. K. Verolme. The SAURON project - II. Sample and early results. MNRAS, 329:513–530, January 2002.

V. P. Debattista, C. M. Carollo, L. Mayer, and B. Moore. Bulges or Bars from Secular Evolution? ApJl, 604:93–96, April 2004.

C. DeGraf, T. Di Matteo, T. Treu, Y. Feng, J.-H. Woo, and D. Park. Scaling relations between black holes and their host galaxies: comparing theoretical and observational measurements, and the impact of selection effects. MNRAS, 454:913–932, November 2015.

P. Di Matteo, F. Combes, I. Chilingarian, A.-L. Melchior, and B. Semelin. Evolution of galaxies in pairs: Learning from simulations. Astronomische Nachrichten, 329:952, December 2008.

X.-B. Dong, L. C. Ho, W. Yuan, T.-G. Wang, X. Fan, H. Zhou, and N. Jiang. A Uniformly Selected Sample of Low-mass Black Holes in Seyfert 1 Galaxies. ApJ, 755: 167, August 2012.

X. Y. Dong and M. M. De Robertis. Low-Luminosity Active Galaxies and Their Central Black Holes. AJ, 131:1236–1252, March 2006. 139 Y. Dubois, C. Pichon, M. Haehnelt, T. Kimm, A. Slyz, J. Devriendt, and D. Pogosyan. Feeding compact bulges and supermassive black holes with low angular momentum cosmic gas at high redshift. MNRAS, 423:3616–3630, July 2012.

A. A. El-Zant, I. Shlosman, M. C. Begelman, and J. Frank. Galaxy Formation in Triaxial Halos: Black Hole-Bulge-Dark Halo Correlation. ApJ, 590:641–653, June 2003.

B. G. Elmegreen and D. M. Elmegreen. Flocculent and grand design spiral galaxies in groups - Time scales for the persistence of grand design spiral structures. ApJ, 267: 31–34, April 1983.

M. Elvis, G. Risaliti, and G. Zamorani. Most Supermassive Black Holes Must Be Rapidly Rotating. ApJl, 565:L75–L77, February 2002.

P. B. Eskridge, J. A. Frogel, R. W. Pogge, A. C. Quillen, R. L. Davies, D. L. DePoy, M. L. Houdashelt, L. E. Kuchinski, S. V. Ram´ırez, K. Sellgren, D. M. Terndrup, and G. P. Tiede. The Frequency of Barred Spiral Galaxies in the Near-Infrared. AJ, 119: 536–544, February 2000.

P. B. Eskridge, J. A. Frogel, R. W. Pogge, A. C. Quillen, A. A. Berlind, R. L. Davies, D. L. DePoy, K. M. Gilbert, M. L. Houdashelt, L. E. Kuchinski, S. V. Ram´ırez, K. Sellgren, A. Stutz, D. M. Terndrup, and G. P. Tiede. Near-Infrared and Optical Morphology of Spiral Galaxies. ApJS, 143:73–111, November 2002.

A. C. Fabian and K. Iwasawa. The mass density in black holes inferred from the X-ray background. MNRAS, 303:L34–L36, February 1999.

N. Fanidakis, C. M. Baugh, A. J. Benson, R. G. Bower, S. Cole, C. Done, and C. S. Frenk. Grand unification of AGN activity in the ΛCDM cosmology. MNRAS, 410: 53–74, January 2011. 140 S. A. Farrell, N. A. Webb, D. Barret, O. Godet, and J. M. Rodrigues. An intermediate- mass black hole of over 500 solar masses in the galaxy ESO243-49. Nature, 460:73–75, July 2009.

L. Ferrarese. Beyond the Bulge: A Fundamental Relation between Supermassive Black Holes and Dark Matter Halos. ApJ, 578:90–97, October 2002.

L. Ferrarese and D. Merritt. A Fundamental Relation between Supermassive Black Holes and Their Host Galaxies. ApJl, 539:L9–L12, August 2000.

F. Ferrari, H. Dottori, N. Caon, A. Nobrega, and D. B. Pavani. The relationship between the S´ersic law profiles measured along the major and minor axes of elliptical galaxies. MNRAS, 347:824–832, January 2004.

A. V. Filippenko and L. C. Ho. A Low-Mass Central Black Hole in the Bulgeless Seyfert 1 Galaxy NGC 4395. ApJl, 588:L13–L16, May 2003.

I. Finkelman and N. Brosch. UGC 4599: a photometric study of the nearest Hoag-type ring galaxy. MNRAS, 413:2621–2632, June 2011.

I. Finkelman, A. Moiseev, N. Brosch, and I. Katkov. Hoag’s Object: evidence for cold accretion on to an elliptical galaxy. MNRAS, 418:1834–1849, December 2011.

D. B. Fisher and N. Drory. Demographics of Bulge Types within 11 Mpc and Implica- tions for Galaxy Evolution. ApJl, 733:L47, June 2011.

L. Foschini, G. Di Cocco, L. C. Ho, L. Bassani, M. Cappi, M. Dadina, F. Gianotti, G. Malaguti, F. Panessa, E. Piconcelli, J. B. Stephen, and M. Trifoglio. XMM- Newton observations of ultraluminous X-ray sources in nearby galaxies. A&A, 392: 817–825, September 2002. 141 M. Freitag, M. A. G¨urkan, and F. A. Rasio. Runaway collisions in young star clusters - II. Numerical results. MNRAS, 368:141–161, May 2006.

D. Friedli and L. Martinet. Bars Within Bars in Lenticular and Spiral Galaxies - a Step in Secular Evolution. A&A, 277:27, September 1993.

M. Fukugita, K. Shimasaku, and T. Ichikawa. Galaxy Colors in Various Photometric Band Systems. PASP, 107:945, October 1995.

M. Fukugita, C. J. Hogan, and P. J. E. Peebles. The Cosmic Baryon Budget. ApJ, 503: 518–530, August 1998.

D. A. Gadotti and R. E. de Souza. NGC 4608 and NGC 5701: Barred Galaxies without Disks? ApJl, 583:75–78, February 2003.

D. A. Gadotti and R. E. de Souza. On the Lengths, Colors, and Ages of 18 Face-on Bars. ApJS, 163:270–281, April 2006.

D. A. Gadotti, E. Athanassoula, L. Carrasco, A. Bosma, R. E. de Souza, and E. Recillas. Near-infrared surface photometry of a sample of barred galaxies. MNRAS, 381:943– 961, November 2007.

J. S. Gallagher, L. S. Sparke, L. D. Matthews, L. M. Frattare, J. English, A. L. Kinney, E. Iodice, and M. Arnaboldi. Wide Field Planetary Camera 2 Observations of the Polar Ring Galaxy NGC 4650A. ApJ, 568:199–209, March 2002.

K. Gebhardt, R. Bender, G. Bower, A. Dressler, S. M. Faber, A. V. Filippenko, R. Green, C. Grillmair, L. C. Ho, J. Kormendy, T. R. Lauer, J. Magorrian, J. Pinkney, D. Rich- stone, and S. Tremaine. A Relationship between Nuclear Black Hole Mass and Galaxy Velocity Dispersion. ApJl, 539:L13–L16, August 2000. 142 K. Gebhardt, R. M. Rich, and L. C. Ho. An Intermediate-Mass Black Hole in the Glob- ular Cluster G1: Improved Significance from New Keck and Hubble Space Telescope Observations. ApJ, 634:1093–1102, December 2005.

K. Gebhardt, J. Adams, D. Richstone, T. R. Lauer, S. M. Faber, K. G¨ultekin, J. Mur- phy, and S. Tremaine. The Black Hole Mass in M87 from Gemini/NIFS Adaptive Optics Observations. ApJ, 729:119, March 2011.

S. Genel, M. Vogelsberger, V. Springel, D. Sijacki, D. Nelson, G. Snyder, V. Rodriguez- Gomez, P. Torrey, and L. Hernquist. Introducing the Illustris project: the evolution of galaxy populations across cosmic time. MNRAS, 445:175–200, November 2014.

J. Gerssen, R. P. van der Marel, K. Gebhardt, P. Guhathakurta, R. C. Peterson, and C. Pryor. Hubble Space Telescope Evidence for an Intermediate-Mass Black Hole in the Globular Cluster M15. II. Kinematic Analysis and Dynamical Modeling. AJ, 124: 3270–3288, December 2002.

R. A. Gonzalez and J. R. Graham. Tracing the Dynamics of Disk Galaxies with Optical and Infrared Surface Photometry: Color Gradients in M99. ApJ, 460:651, April 1996.

A. W. Graham. An Investigation into the Prominence of Spiral Galaxy Bulges. AJ, 121:820–840, February 2001.

A. W. Graham. The black hole mass - spheroid luminosity relation. MNRAS, 379: 711–722, August 2007.

A. W. Graham. Populating the Galaxy Velocity Dispersion - Supermassive Black Hole

Mass Diagram: A Catalogue of (Mbh, σ) Values. PASA, 25:167–175, November 2008.

A. W. Graham. Breaking the Law: The M bh-M spheroid Relations for Core-S´ersic and S´ersic Galaxies. ApJ, 746:113, February 2012. 143 A. W. Graham. Galaxy Bulges and Their Massive Black Holes: A Review. Galactic Bulges, 418:263, 2016.

A. W. Graham and S. P. Driver. A Log-Quadratic Relation for Predicting Supermassive Black Hole Masses from the Host Bulge S´ersic Index. ApJ, 655:77–87, January 2007.

A. W. Graham and N. Scott. The M BH -L spheroid Relation at High and Low Masses, the Quadratic Growth of Black Holes, and Intermediate-mass Black Hole Candidates. ApJ, 764:151, February 2013.

A. W. Graham and N. Scott. The (Black Hole)-bulge Mass Scaling Relation at Low Masses. ApJ, 798:54, January 2015.

A. W. Graham, P. Erwin, N. Caon, and I. Trujillo. A Correlation between Galaxy Light Concentration and Supermassive Black Hole Mass. ApJl, 563:L11–L14, December 2001.

A. W. Graham, S. P. Driver, P. D. Allen, and J. Liske. The Millennium Galaxy Cata- logue: the local supermassive black hole mass function in early- and late-type galaxies. MNRAS, 378:198–210, June 2007.

A. W. Graham, C. A. Onken, E. Athanassoula, and F. Combes. An expanded Mbh- σ diagram, and a new calibration of active galactic nuclei masses. MNRAS, 412: 2211–2228, April 2011.

A. W. Graham, B. C. Ciambur, and R. Soria. Does the Intermediate-Mass Black Hole in LEDA 87300 (RGG 118) Follow the Near-quadratic M M Relation? BH − Spheroid ApJ, 818:172, February 2016.

J. E. Greene and L. C. Ho. Active Galactic Nuclei with Candidate Intermediate-Mass Black Holes. ApJ, 610:722–736, August 2004. 144 J. E. Greene and L. C. Ho. The Mass Function of Active Black Holes in the Local Universe. ApJ, 667:131–148, September 2007.

J. E. Greene, C. Y. Peng, M. Kim, C.-Y. Kuo, J. A. Braatz, C. M. V. Impellizzeri, J. J. Condon, K. Y. Lo, C. Henkel, and M. J. Reid. Precise Black Hole Masses from Megamaser Disks: Black Hole-Bulge Relations at Low Mass. ApJ, 721:26–45, September 2010.

R. D. Grouchy, R. J. Buta, H. Salo, and E. Laurikainen. Ring Star Formation Rates in Barred and Nonbarred Galaxies. AJ, 139:2465–2493, June 2010.

K. G¨ultekin. Astronomy: Intermediate-mass black hole found. Nature, 542:175–176, February 2017.

K. G¨ultekin, D. O. Richstone, K. Gebhardt, T. R. Lauer, S. Tremaine, M. C. Aller, R. Bender, A. Dressler, S. M. Faber, A. V. Filippenko, R. Green, L. C. Ho, J. Ko- rmendy, J. Magorrian, J. Pinkney, and C. Siopis. The M-σ and M-L Relations in Galactic Bulges, and Determinations of Their Intrinsic Scatter. ApJ, 698:198–221, June 2009.

M. A. G¨urkan, M. Freitag, and F. A. Rasio. Formation of Massive Black Holes in Dense Star Clusters. I. Mass Segregation and Core Collapse. ApJ, 604:632–652, April 2004.

M. G. Haehnelt and G. Kauffmann. The correlation between black hole mass and bulge velocity dispersion in hierarchical galaxy formation models. MNRAS, 318:L35–L38, November 2000.

M. G. Haehnelt, P. Natarajan, and M. J. Rees. High-redshift galaxies, their active nuclei and central black holes. MNRAS, 300:817–827, November 1998. 145 N. H¨aring and H.-W. Rix. On the Black Hole Mass-Bulge Mass Relation. ApJl, 604: L89–L92, April 2004.

A. Helmi, J. F. Navarro, A. Meza, M. Steinmetz, and V. R. Eke. On the Nature of the Ringlike Structure in the Outer Galactic Disk. ApJl, 592:L25–L28, July 2003.

M. Hirschmann, K. Dolag, A. Saro, L. Bachmann, S. Borgani, and A. Burkert. Cosmo- logical simulations of black hole growth: AGN luminosities and downsizing. MNRAS, 442:2304–2324, August 2014.

L. C. Ho, Z.-Y. Li, A. J. Barth, M. S. Seigar, and C. Y. Peng. The Carnegie-Irvine Galaxy Survey. I. Overview and Atlas of Optical Images. ApJS, 197:21, December 2011.

A. A. Hoag. A peculiar object in Serpens. AJ, 55:170–170, October 1950.

P. F. Hopkins, L. Hernquist, T. J. Cox, T. Di Matteo, P. Martini, B. Robertson, and V. Springel. Black Holes in Galaxy Mergers: Evolution of Quasars. ApJ, 630:705–715, September 2005.

J. Hu. The black hole mass-stellar velocity dispersion correlation: bulges versus pseudo- bulges. MNRAS, 386:2242–2252, June 2008.

S. Huang, L. C. Ho, C. Y. Peng, Z.-Y. Li, and A. J. Barth. The Carnegie-Irvine Galaxy Survey. III. The Three-component Structure of Nearby Elliptical Galaxies. ApJ, 766: 47, March 2013.

E. Iodice. Narrow Polar Rings versus Wide Polar Ring/Diks Galaxies. In E. Iodice and E. M. Corsini, editors, Multi-Spin Galaxies, ASP Conference Series, volume 486 of Astronomical Society of the Pacific Conference Series, page 39, May 2014. 146 E. Iodice, M. Arnaboldi, L. S. Sparke, J. S. Gallagher, and K. C. Freeman. Near-Infrared photometry in the J, H and Kn bands for Polar Ring Galaxies. I. Data, structural parameters. A&A, 391:103–115, August 2002.

R. I. Jedrzejewski. CCD surface photometry of elliptical galaxies. I - Observations, reduction and results. MNRAS, 226:747–768, June 1987.

L. S. Kelvin, S. P. Driver, A. S. G. Robotham, D. T. Hill, M. Alpaslan, I. K. Baldry, S. P. Bamford, J. Bland-Hawthorn, S. Brough, A. W. Graham, B. H¨aussler, A. M. Hopkins, J. Liske, J. Loveday, P. Norberg, S. Phillipps, C. C. Popescu, M. Prescott, E. N. Taylor, and R. J. Tuffs. Galaxy And Mass Assembly (GAMA): Structural Investigation of Galaxies via Model Analysis. MNRAS, 421:1007–1039, April 2012.

R. Kennedy, S. P. Bamford, I. Baldry, B. H¨außler, B. W. Holwerda, A. M. Hopkins, L. S. Kelvin, R. Lange, A. J. Moffett, C. C. Popescu, E. N. Taylor, R. J. Tuffs, M. Vika, and B. Vulcani. Galaxy And Mass Assembly (GAMA): the wavelength dependence of galaxy structure versus redshift and luminosity. MNRAS, 454:806–817, November 2015.

R. Kennedy, S. P. Bamford, B. H¨außler, I. Baldry, M. Bremer, S. Brough, M. J. I. Brown, S. Driver, K. Duncan, A. W. Graham, B. W. Holwerda, A. M. Hopkins, L. S. Kelvin, R. Lange, S. Phillipps, M. Vika, and B. Vulcani. Galaxy And Mass Assembly (GAMA): understanding the wavelength dependence of galaxy structure with bulge-disc decompositions. MNRAS, 460:3458–3471, August 2016.

N. Khandai, T. Di Matteo, R. Croft, S. Wilkins, Y. Feng, E. Tucker, C. DeGraf, and M.-S. Liu. The MassiveBlack-II simulation: the evolution of haloes and galaxies to z ˜0.MNRAS, 450 : 1349 1374, June2015. −− 147 B. Kızıltan, H. Baumgardt, and A. Loeb. An intermediate-mass black hole in the centre of the globular cluster 47 Tucanae. Nature, 542:203–205, February 2017.

J. H. Knapen, S. Stedman, D. M. Bramich, S. L. Folkes, and T. R. Bradley. Struc- ture and star formation in disk galaxies. II. Optical imaging. A&A, 426:1135–1141, November 2004.

F. Koliopanos, B. C. Ciambur, A. W. Graham, N. A. Webb, M. Coriat, B. Mutlu-Pakdil, B. L. Davis, O. Godet, D. Barret, and M. S. Seigar. Searching for intermediate-mass black holes in galaxies with low-luminosity AGN: a multiple-method approach. A&A, 601:A20, April 2017.

J. Kormendy. The Stellar-Dynamical Search for Supermassive Black Holes in Galactic Nuclei. Coevolution of Black Holes and Galaxies, page 1, 2004.

J. Kormendy and R. Bender. Supermassive black holes do not correlate with dark matter haloes of galaxies. Nature, 469:377–380, January 2011.

J. Kormendy and K. Gebhardt. Supermassive black holes in galactic nuclei. In J. C. Wheeler and H. Martel, editors, 20th Texas Symposium on relativistic astrophysics, volume 586 of American Institute of Physics Conference Series, pages 363–381, Oc- tober 2001.

J. Kormendy and L. C. Ho. Coevolution (Or Not) of Supermassive Black Holes and Host Galaxies. ARA&A, 51:511–653, August 2013.

J. Kormendy and R. C. Kennicutt, Jr. Secular Evolution and the Formation of Pseu- dobulges in Disk Galaxies. ARA&A, 42:603–683, September 2004.

J. Kormendy and D. Richstone. Inward Bound—The Search For Supermassive Black Holes In Galactic Nuclei. ARA&A, 33:581, 1995. 148 J. Kormendy, R. Bender, and M. E. Cornell. Supermassive black holes do not correlate with galaxy disks or pseudobulges. Nature, 469:374–376, January 2011.

K. Kuijken and M. R. Merrifield. Establishing the connection between peanut-shaped bulges and galactic bars. ApJl, 443:L13–L16, April 1995.

A. U. Landolt. UBVRI photometric standard stars in the magnitude range 11.5-16.0 around the celestial equator. AJ, 104:340–371, July 1992.

R. B. Larson. Dynamical models for the formation and evolution of spherical galaxies. MNRAS, 166:585–616, March 1974.

R. B. Larson. Models for the formation of elliptical galaxies. MNRAS, 173:671–699, December 1975.

H. M. Lee. Evolution of galactic nuclei with 10-M black holes. MNRAS, 272:605–617, February 1995.

A. Loeb and F. A. Rasio. Collapse of primordial gas clouds and the formation of quasar black holes. ApJ, 432:52–61, September 1994.

J. M. Lotz, J. Primack, and P. Madau. A New Nonparametric Approach to Galaxy Morphological Classification. AJ, 128:163–182, July 2004.

R. Lynds and A. Toomre. On the interpretation of ring galaxies: the binary ring system II Hz 4. ApJ, 209:382–388, October 1976.

A. V. Macci`o, B. Moore, and J. Stadel. The Origin of Polar Ring Galaxies: Evidence for Galaxy Formation by Cold Accretion. ApJl, 636:L25–L28, January 2006.

P. Madau and M. J. Rees. Massive Black Holes as Population III Remnants. ApJl, 551: L27–L30, April 2001. 149 J. Magorrian, S. Tremaine, D. Richstone, R. Bender, G. Bower, A. Dressler, S. M. Faber, K. Gebhardt, R. Green, C. Grillmair, J. Kormendy, and T. Lauer. The Demography of Massive Dark Objects in Galaxy Centers. AJ, 115:2285–2305, June 1998.

A. Mahdavi and M. J. Geller. A Redshift Survey of Nearby Galaxy Groups: The Shape of the Mass Density Profile. ApJ, 607:202–219, May 2004.

M. Mapelli, B. Moore, E. Ripamonti, L. Mayer, M. Colpi, and L. Giordano. Are ring galaxies the ancestors of giant low surface brightness galaxies? MNRAS, 383:1223– 1231, January 2008.

A. Marconi and L. K. Hunt. The Relation between Black Hole Mass, Bulge Mass, and Near-Infrared Luminosity. ApJl, 589:L21–L24, May 2003.

A. Marconi, G. Risaliti, R. Gilli, L. K. Hunt, R. Maiolino, and M. Salvati. Local supermassive black holes, relics of active galactic nuclei and the X-ray background. MNRAS, 351:169–185, June 2004.

J. W. K. Mark. Studies on the Galactic Density Wave. In J. R. Shakeshaft, editor, The Formation and Dynamics of Galaxies, volume 58 of IAU Symposium, page 417, 1974.

E. E. Mart´ınez-Garc´ıa, I. Puerari, F. F. Rosales-Ortega, R. A. Gonz´alez-L´opezlira, I. Fuentes-Carrera, and A. Luna. The Behavior of the Pitch Angle of Spiral Arms Depending on Optical Wavelength. ApJl, 793:L19, September 2014.

I. Martinez-Valpuesta and I. Shlosman. Why Buckling Stellar Bars Weaken in Disk Galaxies. ApJl, 613:L29–L32, September 2004.

I. Martinez-Valpuesta, I. Shlosman, and C. Heller. Evolution of Stellar Bars in Live Axisymmetric Halos: Recurrent Buckling and Secular Growth. ApJ, 637:214–226, January 2006. 150 F. Marulli, S. Bonoli, E. Branchini, L. Moscardini, and V. Springel. Modelling the cosmological co-evolution of supermassive black holes and galaxies - I. BH scaling relations and the AGN luminosity function. MNRAS, 385:1846–1858, April 2008.

N. J. McConnell and C.-P. Ma. Revisiting the Scaling Relations of Black Hole Masses and Host Galaxy Properties. ApJ, 764:184, February 2013.

N. J. McConnell, C.-P. Ma, J. R. Graham, K. Gebhardt, T. R. Lauer, S. A. Wright, and D. O. Richstone. The Black Hole Mass in the NGC 6086. ApJ, 728:100, February 2011.

M. McDonald, S. Courteau, R. B. Tully, and J. Roediger. A survey of 286 cluster galaxies at optical griz and near-IR H band: surface brightness profiles and bulge-disc decompositions. MNRAS, 414:2055–2068, July 2011.

D. E. McLaughlin, J. Anderson, G. Meylan, K. Gebhardt, C. Pryor, D. Minniti, and S. Phinney. Hubble Space Telescope Proper Motions and Stellar Dynamics in the Core of the Globular Cluster 47 Tucanae. ApJS, 166:249–297, September 2006.

R. J. McLure and J. S. Dunlop. On the black hole-bulge mass relation in active and inactive galaxies. MNRAS, 331:795–804, April 2002.

D. Merritt and L. Ferrarese. Black hole demographics from the M-σ relation. MNRAS, 320:L30–L34, January 2001.

G. E. Miller and J. M. Scalo. The initial mass function and stellar birthrate in the solar neighborhood. ApJS, 41:513–547, November 1979.

M. C. Miller and D. P. Hamilton. Four-Body Effects in Globular Cluster Black Hole Coalescence. ApJ, 576:894–898, September 2002a. 151 M. C. Miller and D. P. Hamilton. Production of intermediate-mass black holes in globular clusters. MNRAS, 330:232–240, February 2002b.

B. Milvang-Jensen and I. Jørgensen. Galaxy Surface Photometry. Baltic Astronomy, 8: 535–574, 1999.

B. Mobasher and P. Mazzei. A Self-Consistent Method for Estimating Photometric Redshifts. In R. Weymann, L. Storrie-Lombardi, M. Sawicki, and R. Brunner, editors, Photometric Redshifts and the Detection of High Redshift Galaxies, volume 191 of Astronomical Society of the Pacific Conference Series, page 37, 1999.

A. V. Moiseev, K. I. Smirnova, A. A. Smirnova, and V. P. Reshetnikov. A new catalogue of polar-ring galaxies selected from the Sloan Digital Sky Survey. MNRAS, 418:244– 257, November 2011.

P. Monaco, P. Salucci, and L. Danese. Joint cosmological formation of QSOs and bulge- dominated galaxies. MNRAS, 311:279–296, January 2000.

E. C. Moran, K. Shahinyan, H. R. Sugarman, D. O. V´elez, and M. Eracleous. Black Holes At the Centers of Nearby Dwarf Galaxies. AJ, 148:136, December 2014.

H. Mouri and Y. Taniguchi. Runaway Merging of Black Holes: Analytical Constraint on the Timescale. ApJl, 566:L17–L20, February 2002.

E. Noyola, K. Gebhardt, and M. Bergmann. Gemini and Hubble Space Telescope Evi- dence for an Intermediate-Mass Black Hole in ω Centauri. ApJ, 676:1008-1015, April 2008.

R. F. Peletier, R. L. Davies, G. D. Illingworth, L. E. Davis, and M. Cawson. CCD surface photometry of galaxies with dynamical data. II - UBR photometry of 39 elliptical galaxies. AJ, 100:1091–1142, October 1990. 152 C. Y. Peng, L. C. Ho, C. D. Impey, and H.-W. Rix. Detailed Decomposition of Galaxy Images. II. Beyond Axisymmetric Models. AJ, 139:2097–2129, June 2010.

D. Pfenniger. Dissolution of a Bar by Merging. In B. Sundelius, editor, Dynamics of Disc Galaxies, page 191, 1991.

Planck Collaboration, P. A. R. Ade, N. Aghanim, C. Armitage-Caplan, M. Arnaud, M. Ashdown, F. Atrio-Barandela, J. Aumont, C. Baccigalupi, A. J. Banday, and et al. Planck 2013 results. XVI. Cosmological parameters. A&A, 571:A16, November 2014.

R. M. Plotkin, S. Markoff, B. C. Kelly, E. K¨ording, and S. F. Anderson. Using the Fundamental Plane of black hole activity to distinguish X-ray processes from weakly accreting black holes. MNRAS, 419:267–286, January 2012.

M. Pohlen and I. Trujillo. The structure of galactic disks. Studying late-type spiral galaxies using SDSS. A&A, 454:759–772, August 2006.

S. F. Portegies Zwart and S. L. W. McMillan. The Runaway Growth of Intermediate- Mass Black Holes in Dense Star Clusters. ApJ, 576:899–907, September 2002.

S. F. Portegies Zwart, J. Dewi, and T. Maccarone. Intermediate mass black holes in accreting binaries: formation, evolution and observational appearance. MNRAS, 355: 413–423, December 2004.

H. Pour-Imani, D. Kennefick, J. Kennefick, B. L. Davis, D. W. Shields, and M. Shameer Abdeen. Strong Evidence for the Density-wave Theory of Spiral Structure in Disk Galaxies. ApJL, 827:L2, August 2016.

I. Puerari and H. A. Dottori. Fourier analysis of structure in spiral galaxies. A&AS, 93: 469–493, June 1992. 153 I. Puerari, D. L. Block, B. G. Elmegreen, J. A. Frogel, and P. B. Eskridge. The detection of spiral arm modulation in the stellar disk of an optically flocculent and an optically grand design galaxy. A&A, 359:932–940, July 2000.

G. D. Quinlan and S. L. Shapiro. The dynamical evolution of dense star clusters in galactic nuclei. ApJ, 356:483–500, June 1990.

P. Rautiainen and H. Salo. N-body simulations of resonance rings in galactic disks. A&A, 362:465–586, October 2000.

A. E. Reines and M. Volonteri. Relations between Central Black Hole Mass and Total Galaxy Stellar Mass in the Local Universe. ApJ, 813:82, November 2015.

V. Reshetnikov, F. Bournaud, F. Combes, M. Fa´undez-Abans, M. de Oliveira-Abans, W. van Driel, and S. E. Schneider. Galaxy transmutations: The double ringed galaxy ESO 474-G26. A&A, 431:503–510, February 2005.

V. P. Reshetnikov, V. A. Hagen-Thorn, and V. A. Yakovleva. Photometric study of polar-ring galaxies. I. UGC 7576 and UGC 9796. A&A, 290:693–704, October 1994.

H.-W. Rix and M. J. Rieke. Tracing the Stellar Mass in M51. ApJ, 418:123, November 1993.

V. C. Rubin. Multi-spin galaxies. AJ, 108:456–467, August 1994.

T. Ryu, T. L. Tanaka, R. Perna, and Z. Haiman. Intermediate-mass black holes from Population III remnants in the first galactic nuclei. MNRAS, 460:4122–4134, August 2016.

P. D. Sackett, H.-W. Rix, B. J. Jarvis, and K. C. Freeman. The flattened dark halo of polar ring galaxy NGC 4650A: A conspiracy of shapes? ApJ, 436:629–641, December 1994. 154 E. E. Salpeter. The Luminosity Function and Stellar Evolution. ApJ, 121:161, January 1955.

P. S´anchez-Bl´azquez, J. Gorgas, N. Cardiel, and J. J. Gonz´alez. Stellar populations of early-type galaxies in different environments. II. Ages and metallicities. A&A, 457: 809–821, October 2006.

A. Sandage and G. A. Tammann. A revised Shapley-Ames Catalog of bright galaxies. 1981.

R. H. Sanders. The effects of dissipation on the gas response to oval distortions of disk galaxies. ApJ, 217:916–927, November 1977.

E. Sani, A. Marconi, L. K. Hunt, and G. Risaliti. The Spitzer/IRAC view of black hole-bulge scaling relations. MNRAS, 413:1479–1494, May 2011.

M. F. Saraiva Schroeder, M. G. Pastoriza, S. O. Kepler, and I. Puerari. The distribution of light in the spiral galaxy NGC 7412. A&AS, 108, November 1994.

G. Savorgnan, A. W. Graham, A. Marconi, E. Sani, L. K. Hunt, M. Vika, and S. P. Driver. The supermassive black hole mass-S´ersic index relations for bulges and ellip- tical galaxies. MNRAS, 434:387–397, September 2013.

G. A. D. Savorgnan. Supermassive Black Holes and their Host Spheroids III. The M n Correlation. ApJ, 821:88, April 2016. bh − sph

G. A. D. Savorgnan, A. W. Graham, A. Marconi, and E. Sani. Supermassive Black

Holes and Their Host Spheroids. II. The Red and Blue Sequence in the M M∗ BH − ,sph Diagram. ApJ, 817:21, January 2016. doi: 10.3847/0004-637X/817/1/21.

J. Schaye, R. A. Crain, R. G. Bower, and M. et al. Furlong. The EAGLE project: 155 simulating the evolution and assembly of galaxies and their environments. MNRAS, 446:521–554, January 2015.

E. F. Schlafly and D. P. Finkbeiner. Measuring Reddening with Sloan Digital Sky Survey Stellar Spectra and Recalibrating SFD. ApJ, 737:103, August 2011.

R. Schneider, A. Ferrara, P. Natarajan, and K. Omukai. First Stars, Very Massive Black Holes, and Metals. ApJ, 571:30–39, May 2002. doi: 10.1086/339917.

R. Sch¨odel, T. Ott, R. Genzel, R. Hofmann, M. Lehnert, A. Eckart, N. Mouawad, T. Alexander, M. J. Reid, R. Lenzen, M. Hartung, F. Lacombe, D. Rouan, E. Gen- dron, G. Rousset, A.-M. Lagrange, W. Brandner, N. Ageorges, C. Lidman, A. F. M. Moorwood, J. Spyromilio, N. Hubin, and K. M. Menten. A star in a 15.2-year orbit around the supermassive black hole at the centre of the Milky Way. Nature, 419: 694–696, October 2002.

A. Schulze and K. Gebhardt. Effect of a Dark Matter Halo on the Determination of Black Hole Masses. ApJ, 729:21, March 2011.

M. P. Schwarz. The response of gas in a galactic disk to bar forcing. ApJ, 247:77–88, July 1981.

M. P. Schwarz. How bar strength and pattern speed affect galactic spiral structure. MNRAS, 209:93–109, July 1984.

F. Schweizer, B. C. Whitmore, and V. C. Rubin. Colliding and merging galaxies. II - S0 galaxies with polar rings. AJ, 88:909–925, July 1983.

F. Schweizer, W. K. Ford, Jr., R. Jedrzejewski, and R. Giovanelli. The structure and evolution of Hoag’s object. ApJ, 320:454–463, September 1987. 156 F. Schweizer, J. H. van Gorkom, and P. Seitzer. The neutral hydrogen ring, mass-to- light ratio, and dark halo of the elliptical galaxy IC 2006. ApJ, 338:770–788, March 1989.

N. Scott, A. W. Graham, and J. Schombert. The Supermassive Black Hole Mass- Spheroid Stellar Mass Relation for S´ersic and Core-S´ersic Galaxies. ApJ, 768:76, May 2013.

M. S. Seigar. The Dark Matter Halo Density Profile, Spiral Arm Morphology, and Supermassive Black Hole Mass of M33. ISRN Astronomy and Astrophysics, 2011: 725697, 2011.

M. S. Seigar and P. A. James. The structure of spiral galaxies - II. Near-infrared properties of spiral arms. MNRAS, 299:685–698, September 1998.

M. S. Seigar, D. L. Block, I. Puerari, N. E. Chorney, and P. A. James. Dust-penetrated arm classes: insights from rising and falling rotation curves. MNRAS, 359:1065–1076, May 2005.

M. S. Seigar, J. S. Bullock, A. J. Barth, and L. C. Ho. Constraining Dark Matter Halo Profiles and Galaxy Formation Models Using Spiral Arm Morphology. I. Method Outline. ApJ, 645:1012–1023, July 2006.

M. S. Seigar, D. Kennefick, J. Kennefick, and C. H. S. Lacy. Discovery of a Relationship between Spiral Arm Morphology and Supermassive Black Hole Mass in Disk Galaxies. ApJl, 678:L93, May 2008.

M. S. Seigar, B. L. Davis, J. Berrier, and D. Kennefick. Constraining Dark Matter Halo Profiles and Galaxy Formation Models Using Spiral Arm Morphology. II. Dark and Stellar Mass Concentrations for 13 Nearby Face-on Galaxies. ApJ, 795:90, November 2014. 157 U. Seljak. Constraints on galaxy halo profiles from galaxy-galaxy lensing and Tully- Fisher/Fundamental Plane relations. MNRAS, 334:797–804, August 2002.

J. A. Sellwood and A. Wilkinson. Dynamics of barred galaxies. Reports on Progress in Physics, 56:173–256, February 1993.

P. Serra, S. C. Trager, J. M. van der Hulst, T. A. Oosterloo, and R. Morganti. VizieR Online Data Catalog: Line-strength indices in IC 4200 (Serra+, 2006). VizieR Online Data Catalog, 345, March 2006.

J. L. Sersic. Atlas de galaxias australes. 1968.

F. Shankar. Black hole demography: from scaling relations to models. Classical and Quantum Gravity, 30(24):244001, December 2013.

F. Shankar, D. H. Weinberg, and J. Miralda-Escud´e. Self-Consistent Models of the AGN and Black Hole Populations: Duty Cycles, Accretion Rates, and the Mean Radiative Efficiency. ApJ, 690:20–41, January 2009.

F. Shankar, D. H. Weinberg, and J. Miralda-Escud´e. Accretion-driven evolution of black holes: Eddington ratios, duty cycles and active galaxy fractions. MNRAS, 428: 421–446, January 2013.

D. W. Shields, B. Boe, C. Pfountz, B. L. Davis, M. Hartley, H. Pour Imani, Z. Slade, D. Kennefick, and J. Kennefick. Spirality: A Novel Way to Measure Spiral Arm Pitch Angle. ArXiv e-prints, November 2015.

D. Sijacki, V. Springel, T. Di Matteo, and L. Hernquist. A unified model for AGN feedback in cosmological simulations of structure formation. MNRAS, 380:877–900, September 2007. 158 D. Sijacki, M. Vogelsberger, S. Genel, V. Springel, P. Torrey, G. F. Snyder, D. Nelson, and L. Hernquist. The Illustris simulation: the evolving population of black holes across cosmic time. MNRAS, 452:575–596, September 2015.

O. K. Sil’chenko and A. V. Moiseev. Nature of Nuclear Rings in Unbarred Galaxies: NGC 7742 and NGC 7217. AJ, 131:1336–1346, March 2006.

J. Silk and M. J. Rees. Quasars and galaxy formation. A&A, 331:L1–L4, March 1998.

O. N. Snaith, B. K. Gibson, C. B. Brook, A. Knebe, R. J. Thacker, T. R. Quinn, F. Governato, and P. B. Tissera. The halo shape and evolution of polar disc galaxies. MNRAS, 425:1967–1979, September 2012.

G. F. Snyder, P. Torrey, J. M. Lotz, S. Genel, C. K. McBride, M. Vogelsberger, A. Pillepich, D. Nelson, L. V. Sales, D. Sijacki, L. Hernquist, and V. Springel. Galaxy morphology and star formation in the Illustris Simulation at z = 0. MNRAS, 454: 1886–1908, December 2015.

A. Soltan. Masses of quasars. MNRAS, 200:115–122, July 1982.

L. S. Sparke. Are polar rings self-gravitating? MNRAS, 219:657–670, April 1986.

M. Spavone, E. Iodice, D. Bettoni, G. Galletta, P. Mazzei, and V. Reshetnikov. A new photometric investigation of the double-ringed galaxy ESO 474-G26: unveiling the formation scenario. MNRAS, 426:2003–2018, November 2012.

M. Spolaor, R. N. Proctor, D. A. Forbes, and W. J. Couch. The Mass-Metallicity Gradient Relation of Early-Type Galaxies. ApJl, 691:L138–L141, February 2009.

V. Springel, T. Di Matteo, and L. Hernquist. Modelling feedback from stars and black holes in galaxy mergers. MNRAS, 361:776–794, August 2005. 159 J. Strader, L. Chomiuk, T. J. Maccarone, J. C. A. Miller-Jones, A. C. Seth, C. O. Heinke, and G. R. Sivakoff. No Evidence for Intermediate-mass Black Holes in Globular Clusters: Strong Constraints from the JVLA. ApJl, 750:L27, May 2012.

C. Struve, T. Oosterloo, R. Sancisi, R. Morganti, and B. H. C. Emonts. Cold gas in massive early-type galaxies: the case of NGC 1167. A&A, 523:A75, November 2010.

A. D. Sutton, T. P. Roberts, and M. J. Middleton. The ultraluminous state revisited: fractional variability and spectral shape as diagnostics of super-Eddington accretion. MNRAS, 435:1758–1775, October 2013.

R. A. Swaters and V. C. Rubin. Stellar Motions in the Polar Ring Galaxy NGC 4650A. ApJl, 587:L23–L26, April 2003.

Y. Taniguchi, K. Shibata, and K.-I. Wakamatsu. New polar ring galaxies in rich clusters of galaxies. Ap&SS, 118:529, January 1986. doi: 10.1007/BF00651181.

V. A. Taylor, R. A. Jansen, R. A. Windhorst, S. C. Odewahn, and J. E. Hibbard. UBVR and Hubble Space Telescope Mid-Ultraviolet and Near-Infrared Surface Photometry and Radial Color Gradients of Late-Type, Irregular, and Peculiar Galaxies. ApJ, 630: 784–803, September 2005.

D. A. Thilker, L. Bianchi, G. Meurer, A. Gil de Paz, S. Boissier, et al. A Search for Extended Ultraviolet Disk (XUV-Disk) Galaxies in the Local Universe. ApJS, 173: 538–571, December 2007.

M. D. Thornley. Uncovering Spiral Structure in Flocculent Galaxies. ApJL, 469:L45, September 1996.

C. E. Thornton, A. J. Barth, L. C. Ho, R. E. Rutledge, and J. E. Greene. The Host 160 Galaxy and Central Engine of the Dwarf POX 52. ApJ, 686: 892-910, October 2008.

D. Tody. The IRAF Data Reduction and Analysis System. In D. L. Crawford, editor, Instrumentation in astronomy VI, volume 627 of Proc. Spie, page 733, January 1986. doi: 10.1117/12.968154.

P. Torrey, S. Wellons, F. Machado, B. Griffen, D. Nelson, V. Rodriguez-Gomez, R. McK- innon, A. Pillepich, C.-P. Ma, M. Vogelsberger, V. Springel, and L. Hernquist. An analysis of the evolving comoving number density of galaxies in hydrodynamical sim- ulations. MNRAS, 454:2770–2786, December 2015.

S. Tremaine, K. Gebhardt, R. Bender, G. Bower, A. Dressler, S. M. Faber, A. V. Filippenko, R. Green, C. Grillmair, L. C. Ho, J. Kormendy, T. R. Lauer, J. Magorrian, J. Pinkney, and D. Richstone. The Slope of the Black Hole Mass versus Velocity Dispersion Correlation. ApJ, 574:740–753, August 2002.

M. Trenti, E. Ardi, S. Mineshige, and P. Hut. Star clusters with primordial binaries - III. Dynamical interaction between binaries and an intermediate-mass black hole. MNRAS, 374:857–866, January 2007.

A. V. Tutukov and A. V. Fedorova. The role of close passages of galaxies and the asymmetry of their dark haloes in the formation of their spiral patterns. Astronomy Reports, 50:785–801, October 2006.

J. S. Ulvestad, J. E. Greene, and L. C. Ho. Radio Emission from the Intermediate-Mass Black Hole in the Globular Cluster G1. ApJl, 661:L151–L154, June 2007.

F. C. van den Bosch, F. Jiang, A. Hearin, D. Campbell, D. Watson, and N. Pad- manabhan. Coming of age in the dark sector: how dark matter haloes grow their gravitational potential wells. MNRAS, 445:1713–1730, December 2014. 161 M. Vika, S. P. Driver, A. W. Graham, and J. Liske. The Millennium Galaxy Catalogue:

the Mbh-Lspheroid derived supermassive black hole mass function. MNRAS, 400:1451– 1460, December 2009.

M. Vika, S. P. Driver, E. Cameron, L. Kelvin, and A. Robotham. The near-IR Mbh-L

and Mbh-n relations. MNRAS, 419:2264–2292, January 2012.

M. Vogelsberger, S. Genel, V. Springel, P. Torrey, D. Sijacki, D. Xu, G. Snyder, D. Nel- son, and L. Hernquist. Introducing the Illustris Project: simulating the coevolution of dark and visible matter in the Universe. MNRAS, 444:1518–1547, October 2014.

M. Volonteri, P. Natarajan, and K. G¨ultekin. How Important is the Dark Matter Halo for Black Hole Growth? ApJ, 737:50, August 2011.

B. Vulcani, S. P. Bamford, B. H¨außler, M. Vika, A. Rojas, N. K. Agius, I. Baldry, A. E. Bauer, M. J. I. Brown, S. Driver, A. W. Graham, L. S. Kelvin, J. Liske, J. Loveday, C. C. Popescu, A. S. G. Robotham, and R. J. Tuffs. Galaxy And Mass Assembly (GAMA): the wavelength-dependent sizes and profiles of galaxies revealed by MegaMorph. MNRAS, 441:1340–1362, June 2014.

K.-I. Wakamatsu. On the nature of Hoag-type galaxy NGC 6028 and related objects. ApJ, 348:448–455, January 1990.

S. D. M. White. Mixing processes in galaxy mergers. MNRAS, 191:1P–4P, April 1980.

B. C. Whitmore, D. B. McElroy, and F. Schweizer. The shape of the dark halo in polar-ring galaxies. ApJ, 314:439–456, March 1987.

B. C. Whitmore, R. A. Lucas, D. B. McElroy, T. Y. Steiman-Cameron, P. D. Sackett, and R. P. Olling. New observations and a photographic atlas of polar-ring galaxies. AJ, 100:1489–1522, November 1990. 162 J. S. B. Wyithe. A log-quadratic relation between the nuclear black hole masses and velocity dispersions of galaxies. MNRAS, 365:1082–1098, February 2006.

D. Zaritsky and K. Y. Lo. Evidence for nonaxisymmetric nuclear bulges in spiral galax- ies. ApJ, 303:66–75, April 1986.

A. V. Zasov, L. N. Petrochenko, and A. M. Cherepashchuk. Supermassive Black Holes: Relation to Dark Halos. Astronomy Reports, 49:362–367, May 2005.

A. Zezas and G. Fabbiano. Chandra Observations of “The Antennae” Galaxies (NGC 4038/4039). IV. The X-Ray Source Luminosity Function and the Nature of Ultralu- minous X-Ray Sources. ApJ, 577:726–737, October 2002. Appendix A

Acronyms

Table A.1: Acronyms

Acronym Meaning

δ Declination ǫ Ellipticity

ǫint Intrinsic scatter

λeff Effective wavelength of a filter µ Mean

µ0 Central surface brightness

µe Effective surface brightness

µB B-band surface brightness

µK K-band surface brightness µm Micron ρ Mass density of BHs

ρ0 Critical mass density of the universe

Continued on next page

163 164 Table A.1 – continued from previous page

Acronym Meaning

ωb Physical baryon density

ΩΛ Dark energy density

ΩM Matter density

ΩBH Cosmological BH mass density σ Standard deviation

σ⋆ Bulge stellar velocity dispersion

σ8 RMS linear fluctuation in the mass distribution on scales of 8/h Mpc τ Time-scale 1D One-dimensional 2D Two-dimensional 2DFFT Two-Dimensional Fast Fourier Transform 2MASS Two Micron All Sky Survey A&A Astronomy and Astrophysics A&AS Astronomy and Astrophysics Supplement AGN Active galactic nuclei AJ Astronomical Journal ALLFITDIST MATLAB code that fits all valid parametric probability distribu- tions to data AM Catalogue of Southern Peculiar Galaxies and Associations ApJ Astrophysical Journal ApJL Astrophysical Journal Letters ApJS Astrophysical Journal Supplement

Continued on next page 165 Table A.1 – continued from previous page

Acronym Meaning

ARA&A Annual Review of Astronomy and Astrophysics arcmin Arcminute arcsec Arcsecond

B-band Blue light wavelength filter with λeff = 445nm b/a Minor- to major-axis ratio B/T Bulge-to-total luminosity BH Super-massive black hole BHMF Black hole mass function BMODEL IRAF task that builds a model image from the results of isophotal analysis CCD Charge-Coupled Device CCDPROC IRAF task that processes mosaic exposures cf. confer, meaning compare with or consult CGS Carnegie-Irvine Galaxy Survey CR Corotation

DL Luminosity (redshift-independent) distance Dec. Declination dex Decimal exponent DAOEDIT IRAF task that edits the DAOPHOT package parameters inter- actively DAOPHOT Dominion Astrophysical Observatory Photometry DM Dark matter E Elliptical

Continued on next page 166 Table A.1 – continued from previous page

Acronym Meaning e.g. exempli gratia, meaning “for example” ELLIPSE IRAF task that fits elliptical isophotes to galaxy images ESA European Space Agency ESO European Southern Observatory et al. et alia, meaning and others F160W Wide Field Camera 3 H wavelength filter on HST FFT Fast Fourier transform FITS Flexible Image Transport System FLATCOMBINE IRAF task that combines and processes mosaic flat field exposures FMEDIAN IRAF fast median algorithm F P BH Fundamental plane of black-hole activity − FWHM Full width at half maximum G Gravitational constant (in Chapter 2) G Gini coefficient (in Chapter 3) GALEX Galaxy Evolution Explorer GALFIT 2D galaxy fitting algorithm written in C GAUSS IRAF task to convolve image with a Gaussian function GIM2D Galaxy Image 2D, which is an IRAF package written to perform detailed bulge/disk decompositions Gyr Gigayear

H-band NIR wavelength filter with λeff = 1.65µm

H0 Hubble constant

HI Neutral atomic hydrogen gas

Continued on next page 167 Table A.1 – continued from previous page

Acronym Meaning

HLXs Hyper-luminous X-ray sources HST Hubble Space Telescope HYPERZ Photometric redshift code

I-band Infrared light wavelength filter with λeff = 806nm IC Index Catalogue i.e. id est, meaning “that is” IDL Interactive Data Language IGM Intergalactic medium ILR Inner Lindblad resonance IMBH Intermediate-mass black hole IMCNTR IRAF image center routine IMCOMBINE IRAF task that combines images using various algorithms IMF Initial mass function IMREPLACE IRAF task that replaces a range of pixel values with a constant IMSTAT IRAF task that computes and prints image pixel statistics IMTRANSPOSE IRAF task that transposes a list of 2D images IPAC Infrared Processing and Analysis Center IRAF Image Reduction and Analysis Facility Irr Irregular

J-band NIR wavelength filter with λeff = 1.25µm

K-band NIR wavelength filter with λeff = 2.17µm km Kilometer

LBulge Bulge near-infrared luminosity

Continued on next page 168 Table A.1 – continued from previous page

Acronym Meaning

LCO Las Campanas Observatory LF Luminosity function LLAGN Low-luminosity active galactic nuclei log Base 10 logarithm m Meter m Harmonic mode in the 2DFFT algorithm mB Apparent B-band magnitude mV Apparent V-band magnitude mJ Apparent J-band magnitude mtotal Total mtotalE Total apparent magnitude derived from ellipses mtotalC Total apparent magnitude derived from circles

MB Absolute B-band magnitude

Msph Spheroid absolute magnitude

MV Absolute V-band magnitude

M20 Second-order moment of the brightest 20% of the galaxy’s flux

M⊙ Solar mass

MBH BH mass

MDM Dark matter mass

Mhalo Dark matter halo mass

M⋆,2R Stellar mass within the 2x stellar half-mass radius

M⋆,Bulge Bulge stellar mass

M⋆,sph Bulge stellar mass

Continued on next page 169 Table A.1 – continued from previous page

Acronym Meaning

M⋆,total Total stellar mass mag Magnitude MAST Mikulski Archive for Space Telescopes max Maximum MCMC Markov Chain Monte Carlo MGC Millennium Galaxy Catalogue MIR Mid-infrared MNRAS Monthly Notices of the Royal Astronomical Society Mpc Megaparsec MPFIT IDL Curve Fitting routine Myr Megayear n S´ersic index NASA National Aeronautics and Space Administration NED NASA/IPAC Extragalactic Database NGC of Nebulae and Clusters of Stars NIR Near-infrared nm Nanometer NORMRND Random number generator in MATLAB NSF National Science Foundation NUV Near-ultraviolet OLR Outer Lindblad resonance P Spiral arm pitch angle

PB B-band Spiral arm pitch angle

Continued on next page 170 Table A.1 – continued from previous page

Acronym Meaning

PR R-band Spiral arm pitch angle

PK K-band Spiral arm pitch angle P2DFFT Parallelized 2D Fast Fourier Transform routine P.A. Position Angle, measured counterclockwise from the +y-axis PASA Publications of the Astronomical Society of Australia PASP Publications of the Astronomical Society of the Pacific pc PDF Probability Density Function PEARSPDF MATLAB code that returns the probability distribution denisty of the pearsons distribution with mean (µ), standard deviation σ, skewness and kurtosis. PGC Principle Galaxies Catalogue PRG Polar ring galaxy PSF Point Spread Function RA Right Ascension

Rdisk Disk scale length

Re Bulge effective radius

R-band Red light wavelength filter with λeff = 658nm RC3 Third Reference Catalogue of Bright Galaxies RMS Root mean square RMSE Root mean square error s Second s-BH Stellar-mass black hole

Continued on next page 171 Table A.1 – continued from previous page

Acronym Meaning

S0 Lenticular SDSS Sloan Digital Sky Survey SED Spectral energy distribution SExtractor Source Extractor SFH Star formation history SFR Star formation rate SMA Semimajor axis stdev Standard deviation (σ) STScI Space Telescope Science Institute STSDAS Space Telescope Science Data Analysis System tL Lookback time UGC Uppsala General Catalogue ULXs Ultra-luminous X-ray sources vc Circular velocity vr Radial velocity vvir Virial velocity

Vc Comoving volume

V-band Visible light wavelength filter with λeff = 551nm

W1 MIR wavelength filter with λeff = 3.4µm on WISE

W2 MIR wavelength filter λeff = 4.6µm on WISE

W3 MIR wavelength filter λeff = 12µm on WISE

W4 MIR wavelength filter λeff = 22µm on WISE WISE Wide-field Infrared Survey Explorer

Continued on next page 172 Table A.1 – continued from previous page

Acronym Meaning

WMAP Wilkinson Microwave Anisotropy Probe z Redshift Z Metallicity

Z⊙ Solar metallicity ZEROCOMBINE IRAF task that combines and processes mosaic zero level expo- sures Appendix B

University of Minnesota Duluth News Release on PGC 1000714

Researchers get first look at new and extremely rare galaxy

Approximately 359 million light-years away from Earth, there is a galaxy with an in- nocuous name (PGC 1000714) that doesnt look quite like anything astronomers have observed before. New research provides a first description of a well-defined elliptical- like core surrounded by two circular rings, a galaxy that appears to belong to a class of galaxies rarely observed, Hoag-type galaxies. This work was done by scientists at the University of Minnesota Duluth and the North Carolina Museum of Natural Sciences. (See video about this research here: https://youtu.be/F5ds04xtRdY). Less than 0.1% of all observed galaxies are Hoag-type galaxies, says Bur¸cin Mutlu- Pakdil, lead author of a paper on this work and a graduate student at the University of Minnesota Duluth, Minnesota Institute for Astrophysics and University of Minnesota Twin Cities. Hoag-type galaxies are round cores surrounded by a circular ring, with nothing visibly connecting them. The majority of observed galaxies are disc-shaped

173 174 like our own Milky Way. Galaxies with unusual appearances give astronomers unique insights into how galaxies are formed and change. The researchers collected multi-waveband images of the galaxy, which is only easily observable in the southern hemisphere, using a large diameter telescope in the Chilean mountains. These images were used to determine the ages of the two main features of the galaxy, the outer ring and the central body. While the researchers found a blue and young (0.13 billion years) outer ring, sur- rounding a red and older (5.5 billion years) central core, they were surprised to uncover evidence for second inner ring around the central body. To document this second ring, researchers took their images and subtracted out a model of the core. This allowed them to observe and measure the obscured, second inner ring structure. Weve observed galaxies with a blue ring around a central red body before, the most well-known of these is Hoags object. However, the unique feature of this galaxy is what appears to be an older diffuse red inner ring,, says Patrick Treuthardt, co-author of the study and an astrophysicist at the North Carolina Museum of Natural Sciences. Galaxy rings are regions where stars have formed from colliding gas. The different colors of the inner and outer ring suggests that this galaxy has experienced two different formation periods, Mutlu-Pakdil says. From these initial single snapshots in time, its impossible to know how the rings of this particular galaxy were formed. The researchers say that by accumulating snapshot views of other galaxies like this one astronomers can begin to understand how unusual galaxies are formed and evolve. While galaxy shapes can be the product of internal or external environmental in- teractions, the authors speculate that the outer ring may be the result of this galaxy incorporating portions of a once nearby gas-rich dwarf galaxy. They also say that infer- ring the history of the older inner ring would require the collection of of higher-resolution infrared data. 175 Whenever we find a unique or strange object to study, it challenges our current theories and assumptions about how the Universe works. It usually tells us that we still have a lot to learn, says Treuthardt.

###

An uncorrected proof of the peer-reviewed paper, A photometric study of the peculiar and potentially double ringed, nonbarred galaxy: PGC 1000714, was published online in the Monthly Notices of the Royal Astronomical Society by Oxford University Press on Nov. 30, 2016 and will appear online at http://doi.org/10.1093/mnras/stw3107. The paper was co-authored by Mithila Mangedarage and Marc S. Seigar of the University of Minnesota Duluth. Bur¸cin Mutlu-Pakdil received funding for this work from UMDs Swenson College of Science and Engineering. Computational equipment was provided by the Theodore Dunham, Jr. Fund for Astrophysical Research. The study abstract follows.

A photometric study of the peculiar and potentially double ringed, nonbarred galaxy: PGC 1000714

Authors: Bur¸cin Mutlu-Pakdil, University of Minnesota Duluth and University of Min- nesota Twin Cities; Mithila Mangedarage, University of Minnesota Duluth; Marc S. Seigar, University of Minnesota Duluth; Patrick Treuthardt, North Carolina Museum of Natural Sciences Abstract: We present a photometric study of PGC 1000714, a galaxy resembling Hoag’s Object with a complete detached outer ring, that has not yet been described in the literature. Since the Hoag-type galaxies are extremely rare and peculiar systems, it is necessary to increase the sample of known objects by performing the detailed 176 studies on the possible candidates to derive conclusions about their nature, evolution, and systematic properties. We therefore performed surface photometry of the central body by using the archival near-UV, infrared data and the new optical data (BVRI). This current work has revealed for the first time an elliptical galaxy with two fairly round rings. The central body follows well a r1/4 light profile, with no sign of a bar or stellar disc. By reconstructing the observed spectral energy distribution (SED), we recover the stellar population properties of the central body and the outer ring. Our work suggests different formation histories for the galaxy components. Possible origins of the galaxy are discussed, and we conclude that a recent accretion event is the most plausible scenario that accounts for the observational characteristic of PGC 1000714.