ABSTRACT
Title of dissertation: FROM TANTRUMS
TO TRANSFORMATIONS:
AGN TRANSIENTS DISCOVERED WITH ZTF
Sara Evalyn Frederick
Doctor of Philosophy, 2021
Dissertation directed by: Dr. Suvi Gezari
Space Telescope Science Institute, Maryland
This dissertation work has consisted of searches for extreme AGN-related out-
bursts during Phase I of the Zwicky Transient Facility (ZTF) survey, which has
been a ground-breaking wide-field instrument for the real-time detection and reg-
ular cadence monitoring of transients in the Northern Sky. Transients found to
be nuclear through photometric filtering were vetted by humans and coordinated
for prompt follow-up with various rapid robotic, spectroscopic, and high energy re-
sources, to understand the nature of the galaxy centers undergoing flares and the
appearance of spectral features. Findings from this unprecedentedly high-volume
data stream were often serendipitous, and led to surprising new avenues for study,
including 1) the establishment of a new observational class of quiescent galaxies
caught turning into quasars, 2) the discovery of a preponderance of smooth and
high-amplitude optical transients hosted in NLSy1s, and 3) a framework for distin-
guishing extreme AGN variability from other transients in AGN. We present the
results of these observations, including candidates for TDEs in AGN, changing-look
AGN caught ?turning-on?, as well as members of the new emerging observational
class of flares in Narrow-Line Seyfert 1 (NLSy1) galaxies associated with enhanced
accretion (Trakhtenbrot et al. 2019). We compared the properties of these samples
of flares to previously reported changing-look quasars and Seyfert galaxies, con-
firmed that they are a unique observational class of transients related to physical
processes associated with the central supermassive black hole?s accretion state, and
considered the observations in the context of the physical interpretations for a range
of related transients from the literature. With these unique sample sets, we also aim
to understand why we have found certain galaxy types to preferentially host the sites
of such rapid enhanced flaring activity, and attempt to map out the innermost en-
vironment of the accretion events. These pathfinding studies enabled with ZTF
have the potential to guide how these exceptional moments of AGN evolution will
be systematically discovered in future large area surveys such as the Vera C. Rubin
Observatory.
FROM TANTRUMS TO TRANSFORMATIONS:
AGN Transients
Discovered with
the Zwicky Transient Facility
by
Sara Evalyn Frederick
Dissertation submitted to the Faculty of the Graduate School of the
University of Maryland, College Park in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
2021
Advisory Committee:
Professor Suvi Gezari, Chair and Research Advisor
Professor Richard Mushotzky
Professor Sylvain Veilleux
Dr. Stephanie LaMassa
Professor Kara Hoffman, Dean?s Representative
? Copyright by
Sara Evalyn Frederick
2021

Preface
A portion of this dissertation was previously published as first-author peer-
reviewed journal articles and presented at professional conferences and collaboration
meetings, with contributions from co-authors, builders, and members of the Black
Holes Science Working Group of the ZTF collaboration who made this work possible,
including (but not limited to) Matthew Graham, Shri Kulkarni, Dan Stern, Sjoert
van Velzen, Jesper Sollerman, Daniel Perley, Tiara Hung, Charlotte Ward, and many
others.
In order of publication:
Chapter 2 has been published in the Astrophysical Journal, Volume 883, Issue
1 as ?A New Class of Changing-Look LINERS.?
Chapter 3 has been accepted for publication in the Astrophysical Journal as
?A Family Tree of Optical Transients in Narrow-line Seyfert 1 Galaxies?.
Preliminary results of Chapter 4 have been presented at the Joint Space-
Science Institute (JSI1) Fall 2019 Meeting in Annapolis, Maryland, the Spring 2020
ZTF Virtual Collaboration Meeting hosted by Humboldt-Universita?t zu Berlin, and
the Winter 2021 237th ?Virtually Anywhere? American Astronomical Society Meet-
ing. It will be submitted for publication in the Astrophysical Journal in the Fall of
2021.
1https://jsi.astro.umd.edu/
ii
Dedication
To my loving and proud parents, to my grandparents for modeling
meaningful life-long hard work in pursuit of scholarship, and to
scholars-in-training who aspire to pave divergent paths while lifting
others as you climb: You are our universe?s past and future
ancestors, and this science is for you.
iii
Acknowledgments
The recipe for my success included countless benefactors and support struc-
tures sometimes but not always baked into the systems and institutions I navigated.
Thank you so much, to all of you, for seeing something in my journey worth attend-
ing to as I figured out my path.
My parents Maria Pineda and Steve Frederick for believing in me, for the care
packages, for editing my draft text, and for supporting my academic journey even
when it kept us apart. And to my extended family, especially my grandparents Julio
and Elia Pineda and Harold and Thelma Frederick, for my mind, storytelling skills,
culture, bloodline, and your unconditional pride (even calling me ?Doctora? before
I had the papers signed). Even from a distance, you cured me when I needed family,
community, and love.
My various formal and informal advisors, especially Suvi Gezari, Erin Kara,
Stephen Privitera, and Sjoert van Velzen, for positive examples, for your patience
and guidance, as well as professional networking and development resources. I can?t
thank you enough for your time, skills, code, enthusiasm, advice, and persistent
attention. To the physicists like me I could look up to, the LUMA and MITES
programs and especially Prof. Peter Gonthier, for taking a chance on the one external
student who applied to your research program, for taking me under your wing and
making me feel like I belonged in your home, your lab, at NASA Goddard, and in
the field of astrophysics. And to the ZTF collaboration team, especially Thomas
Kupfer, Umma Rebbapragada, Matthew Graham, Shri Kulkarni, and Roger Smith,
iv
for facilitating training, funding, experiences, and friendships that made me a better,
more confident, and connected scientist.
All the supportive, curative counter-spaces I stumbled into, including but not
limited to MICA, LASC, MBSA, PAARC, CMRS, the now dismantled MD Food Co-
op, and to the navigators who welcomed me graciously into these spaces, especially
to Lisa Warren, Ghonva Ghauri, Dorothy Kou, Dr. Naliyah Kaya and for the many
discussions on identity, indigeneity, art, social justice, academia, and activism that
sustained and nourished me throughout my time on the UMD campus.
To all the change leaders I had the honor to overlap with within the Maryland
Astronomy Department, who strove every day to prioritize the success of every
scientist, including but certainly not limited to Amy Steele, Laura Blecha, Alice
Olmstead, Katie Jameson, Ashley Wilkins, Stuart Vogel, Barbara Hansborough,
Derek Richardson, Krista Smith, Brian Morsony, Taro Shimizu, and many, many
more. My personal and professional growth was directly related to actions you
had the vision to prioritize, personal privileges you fought to dismantle, examples
and norms you set, and doors you opened so I could walk through. I saw and
benefited from your efforts, and it was truly an honor to look up to you all as role
models in the educational, social justice, and research spaces in which we co-existed.
Thank you for being gracious friends, as well as my role models and career mentors.
Additionally, to the business and administrative staff at the University of Maryland
for your patience with and kindness toward me, and for upholding and improving
the department, college, and university through your vital work, especially Dorinda
Kimbrall, Natalie Rowe, Susan Lehr, John Cullinan, and Adrienne Newman.
v
All the unofficial host families who took me in, stored my belongings, shared
hand-me-down housewares and life advice, invited me to hot meals and religious
services on holidays, and made sure I never went hungry or lonely in between,
especially the Fosters, the Louies, the Ritonias, and the Gershteyns. Thank you
for welcoming me into your lives, and making my time away from home much more
enjoyable.
My tabletop gaming families, especially game runners Elece Smith, Zak Glen-
non, Teal, and Joe DeMartini: for hosting inclusive campaigns and for the long
nights we collectively clung to normalcy during the Covid-19 pandemic.
Pooja Suresh for being a sister I never had, for lending me Physics GRE study
materials and passing along advice from your PhD-educated Dad, and for giving me
a brilliant physicist to look up to and to talk through milestones for hours as we
went on this grad school journey together. Where would I be without you?
Finally, acknowledgements would not be complete without expressing grati-
tude to the sanitation and maintenance staff in my various workplaces for enabling
this work, and last but certainly not least to the original stewards of the various
dispossessed locations which I occupied while completing this dissertation, for pre-
serving and defending these places for generations, including but not limited to the
Maya, Sami, Gaels, Cheyenne, Oglala Sioux, Arapaho, Piscatawey, Pamunkey, Hopi,
Pueblo, Apache, Duwamish, Tongva, Carib, and Taino people of the contemporary
and ancestral land now called Honduras, Sweden, Scotland, and the United States of
(Northern) America, also known as Turtle Island, in states and territories including
land that is known as Nebraska, Maryland, Washington D.C., Arizona, Washington,
vi
California, and the Virgin Islands.2
Thank you all.
2https://native-land.ca/
vii
Table of Contents
Preface ii
Dedication iii
Acknowledgements iv
List of Tables xii
List of Figures xiii
List of Abbreviations xv
1 Introduction 1
1.1 Introduction to SMBHs and AGN . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Anatomy of an AGN . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1.1 The X-ray Emitting Corona . . . . . . . . . . . . . . 4
1.1.1.2 The Soft X-ray Excess . . . . . . . . . . . . . . . . . 5
1.1.1.3 Gas Clouds and Obscuring Structures in AGN . . . . 5
1.1.2 AGN Variability . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 Changing-look AGN: A Unique Challenge to AGN Theory . . . . . . 7
1.2.1 The AGN Unification Scheme and Select Host Galaxy Types . 7
1.2.2 Intermediate AGN Type Classifications . . . . . . . . . . . . . 9
1.2.3 Narrow Line Seyfert 1 Galaxies . . . . . . . . . . . . . . . . . 10
1.2.4 The Heterogeneous Population of Low Ionization Nuclear Emis-
sion Line Regions (LINERs) . . . . . . . . . . . . . . . . . . . 11
1.2.5 Review of Changing-Look AGN . . . . . . . . . . . . . . . . . 13
1.3 Overview of Thesis Work . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.3.1 Why Study AGN with ZTF? . . . . . . . . . . . . . . . . . . . 18
1.3.2 Summary of Chapters . . . . . . . . . . . . . . . . . . . . . . 19
viii
2 A New Class of Changing-Look LINERs 23
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2 Discovery and Observations . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.1 Sample Selection Criteria . . . . . . . . . . . . . . . . . . . . . 26
2.2.2 ZTF Light Curve . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.3 Capturing the Transition in Archival Light Curves . . . . . . 30
2.2.4 Host Galaxy Morphology . . . . . . . . . . . . . . . . . . . . . 33
2.2.5 Optical Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.6 UV Imaging and Spectroscopy . . . . . . . . . . . . . . . . . . 39
2.2.7 X-ray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.2.8 Infrared . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.2.9 Radio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.3 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.3.1 Host Galaxy Classification . . . . . . . . . . . . . . . . . . . . 47
2.3.2 Black Hole Masses . . . . . . . . . . . . . . . . . . . . . . . . 50
2.3.3 Comparison to Tidal Disruption Events . . . . . . . . . . . . . 51
2.3.4 Comparison to Seyfert CLAGN . . . . . . . . . . . . . . . . . 53
2.3.5 Eddington Ratio Estimates . . . . . . . . . . . . . . . . . . . 55
2.3.6 ZTF18aajupnt: A LINER Changing-Look to a NLS1 . . . . . 58
2.3.6.1 Coronal Line Emission from ZTF18aajupnt . . . . . 59
2.3.6.2 ZTF18aajupnt as a NLS1 in its ?On? State . . . . . 61
2.3.6.3 The Accretion Rate of ZTF18aajupnt . . . . . . . . 64
2.3.6.4 X-ray Light Curve and Spectra of ZTF18aajupnt . . 65
2.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.4.1 A New Class of changing-look LINERs . . . . . . . . . . . . . 68
2.4.2 Is ZTF18aajupnt a TDE or AGN activity? . . . . . . . . . . . 68
2.4.3 The nature of the high-ionization forbidden ?coronal? lines in
ZTF18aajupnt . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.4.4 The nature of the soft X-ray excess during the NLS1 state of
ZTF18aajupnt . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.6 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3 A Family Tree of Optical Transients from Narrow-Line Seyfert 1 Galaxies 80
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.2 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3.2.1 Optical Photometry . . . . . . . . . . . . . . . . . . . . . . . . 84
3.2.2 Optical Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . 86
3.2.3 UV Photometry . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.2.4 X-rays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.2.5 IR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.3 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.3.1 Photometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.3.1.1 Light Curve Timescales . . . . . . . . . . . . . . . . 93
3.3.1.2 Rebrightening . . . . . . . . . . . . . . . . . . . . . . 93
ix
3.3.1.3 UV/Optical to X-ray Ratio . . . . . . . . . . . . . . 94
3.3.2 Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.3.2.1 Strong He II profiles in AGN? . . . . . . . . . . . . . 95
3.3.2.2 The Fe II complex . . . . . . . . . . . . . . . . . . . 96
3.3.3 X-rays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.3.4 Black Hole Masses . . . . . . . . . . . . . . . . . . . . . . . . 97
3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.4.1 ?IIn or not IIn??: Preliminary Observational Classification of
the Flare Sample . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.4.2 A Preponderance of Rapid Optical Transients in Narrow-line
Seyfert 1 Host Galaxies . . . . . . . . . . . . . . . . . . . . . . 100
3.4.3 Observational Classification: The ?Family Tree? of NLSy1-
associated Transients . . . . . . . . . . . . . . . . . . . . . . . 103
3.4.4 Physical Interpretation of the Transient Flares . . . . . . . . . 104
3.4.4.1 Association of the Transients with AGN . . . . . . . 105
3.4.4.2 The SN Scenario . . . . . . . . . . . . . . . . . . . . 106
3.4.4.3 The TDE Scenario . . . . . . . . . . . . . . . . . . . 107
3.4.4.4 The Extreme AGN Variability Scenario . . . . . . . 109
3.4.4.5 The Gravitational Microlensing Scenario . . . . . . . 110
3.4.4.6 The SMBH Binary Scenario . . . . . . . . . . . . . . 111
3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.6 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4 The X-ray View of a New Class of Changing-look LINERs 124
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
4.2 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.2.1 Target Selection . . . . . . . . . . . . . . . . . . . . . . . . . . 128
4.2.2 Archival Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
4.3 Optical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
4.3.1 ZTF Photometry . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.3.1.1 Optical Monitor . . . . . . . . . . . . . . . . . . . . 133
4.3.2 Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
4.3.3 X-rays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
4.3.3.1 Swift . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
4.3.3.2 XMM . . . . . . . . . . . . . . . . . . . . . . . . . . 136
4.3.3.3 NuSTAR . . . . . . . . . . . . . . . . . . . . . . . . 137
4.4 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
4.4.1 X-rays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
4.4.2 Optical Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . 147
4.4.3 Black Hole Mass Estimates . . . . . . . . . . . . . . . . . . . . 151
4.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
4.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
4.7 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
x
5 Conclusions and Future Work 161
5.1 Lessons Learned . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
5.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
5.2.1 Looking Forward to the Landscape of Real-time All-sky and
High Energy Observations . . . . . . . . . . . . . . . . . . . . 163
5.3 Overview of Contributions . . . . . . . . . . . . . . . . . . . . . . . . 165
A Facilities and Software 168
xi
List of Tables
1 Abbreviations Used In This Thesis . . . . . . . . . . . . . . . . . . . xv
2.1 Summary of Changing-look LINER ZTF Year 1 Sample Data . . . . 28
2.2 LINER Host Galaxy Properties . . . . . . . . . . . . . . . . . . . . . 35
2.3 Spectroscopic Follow-up Observations of ZTF18aajupnt . . . . . . . . 37
2.4 Coronal Line Measurements of ZTF18aajupnt . . . . . . . . . . . . . 40
2.5 Swift Photometry of ZTF18aajupnt . . . . . . . . . . . . . . . . . . . 44
3.1 Spectroscopic Follow-Up of NLSy1 Transients Sample . . . . . . . . . 87
3.2 Black Hole Mass Measurements of NLSy1 Transient Sample . . . . . 98
3.3 Observed Properties and Classifications of NLSy1 Sample and Re-
lated Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4.1 Archival X-ray Observations of CL-LINERs . . . . . . . . . . . . . . 129
4.2 Optical Properties of Full Changing-look LINER ZTF Phase I Sample 131
4.3 NuSTAR Observations of CL LINER candidates . . . . . . . . . . . 139
4.4 Swift XRT Observations of CL LINER Candidates . . . . . . . . . . 140
4.5 XMM EPIC pn Observations of CL LINER candidates . . . . . . . . 141
4.6 Black Hole Mass Estimates of ZTF Year 3 CL-LINER Sample . . . . 152
xii
List of Figures
1.1 XRB Analogy with Changing-look AGN . . . . . . . . . . . . . . . . 3
1.2 Unification Model of Active Galactic Nuclei . . . . . . . . . . . . . . 8
1.3 BPT Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4 Elitzur et al. 2014 Evolutionary Sequence of AGN Intermediate Types 15
1.5 iPTF 16bco Spectra Before and After Changing-look Transition . . . 16
1.6 Physical Explanation for Rapid Changing-look Quasar Transition
Timescales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 Changing-look LINER Year 1 Sample Light Curves . . . . . . . . . . 29
2.2 Long-term Host Photometry of Year 1 CL-LINER Sample . . . . . . 31
2.3 Optical Images of CL-LINER Host Galaxies . . . . . . . . . . . . . . 34
2.4 ZTF Year 1 CL-LINER Sample Spectra . . . . . . . . . . . . . . . . . 36
2.5 Shapes of Best-Sampled ZTF Light Curves of the CL LINER Year 1
Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.6 Comparison of ZTF18aajupnt Spectra with Coronal Line Emitters . . 39
2.7 Coronal Lines in Keck Spectrum of ZTF18aajupnt . . . . . . . . . . . 40
2.8 Multi-wavelength Difference Imaging Light Curve of ZTF18aajupnt . 41
2.9 UV Spectrum of ZTF18aajupnt . . . . . . . . . . . . . . . . . . . . . 42
2.10 Swift X-ray Follow-up of ZTF18aajupnt . . . . . . . . . . . . . . . . . 45
2.11 XMM-Newton Follow-up of ZTF18aajupnt . . . . . . . . . . . . . . . 45
2.12 Host Galaxy PPXF Fits of CL-LINER Year 1 Sample . . . . . . . . . . 48
2.13 BPT Diagram of CL LINER Year 1 Sample . . . . . . . . . . . . . . 49
2.14 Flux Change of CL LINER Sample compared to CLQs . . . . . . . . 54
2.15 Elitzur et al. 2014 Evolutionary Sequence of AGN Types . . . . . . . 57
2.16 Line Widths in ZTF18aajupnt Optical Follow-up Spectra . . . . . . . 61
2.17 Emission Line Fits for CL-LINER Year 1 Sample Compared to CLQs 63
2.18 Hydrogen and Oxygen Emission Line Fits for CL-LINER Year 1 Sample 76
2.19 Line Fits to Low Ionization Emission Lines of CL LINER Year 1 Sample 77
2.20 Changing-Look LINER Year 1 Sample Follow-up Line Fits . . . . . . 78
2.21 Coronal Line Fits to ZTF18aajupnt Follow-up Spectrum . . . . . . . 79
3.1 ZTF Difference Imaging Light Curves of NLSy1 Transient Sample . . 85
3.2 Multiwavelength Light Curves of ZTF Sample of NLSy1 Transients . 92
xiii
3.3 Gaussian Fits to Optical Light Curves of NLSy1 Transients . . . . . . 112
3.4 Rise Times vs. Absolute Magnitude of NLSy1 Transients and Related
Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
3.5 Balmer and He II Line Profiles of NLSy1 Transient Sample . . . . . . 114
3.6 Spectra of NLSy1 Transients and Related Events from the Literature 115
3.7 Optical Spectroscopic Features of NLSy1 Transients and Related Events116
3.8 X-ray Spectroscopy of AT2019pev and AT2019avd . . . . . . . . . . . 117
3.9 Comparison of NLSy1 Spectra to NLSy1, SN, & TDE . . . . . . . . . 118
3.10 ZTF Forced Photometry of NLSy1 Transients . . . . . . . . . . . . . 122
3.11 All Spectroscopic Follow-up of NLSy1 Optical Transients . . . . . . . 123
4.1 ZTF Light Curves of Full CL-LINER Sample . . . . . . . . . . . . . . 132
4.2 ZTF Forced Photometry of ZTF18aahmkac, ZTF19aambzmf, and
ZTF18aaavffc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
4.3 Discovery Spectra of ZTF Year 2-3 CL-LINER Sample . . . . . . . . 135
4.4 XMM-Newton Images of CL-LINERs . . . . . . . . . . . . . . . . . . 138
4.5 NuSTAR Images of CL-LINERs . . . . . . . . . . . . . . . . . . . . . 139
4.6 XMM-Newton Spectra of CL-LINERs . . . . . . . . . . . . . . . . . . 142
4.7 NuSTAR Spectra of CL-LINERs . . . . . . . . . . . . . . . . . . . . . 143
4.8 Comparison of CL-LINERs to BLAGN and CLQ Samples in ?OX and
Eddington Ratio Parameter Space . . . . . . . . . . . . . . . . . . . . 144
4.9 X-ray Flux Densities vs. Eddington Ratios of CL-LINER Sample . . 145
4.10 Evolution of ?OX of CL-LINER Sample . . . . . . . . . . . . . . . . . 146
4.11 Follow-up Multi-epoch Spectroscopy of ZTF Year 3 CL-LINER Sample148
4.12 Flux Change of ZTF Year 3 CL LINER Sample Compared to CLQs . 150
4.13 H? vs. Oxygen Line Luminosities for ZTF Year 3 CL-LINER Sample 151
4.14 Model fits to Balmer and Oxygen Emission Lines of Latest CL-LINERs
in the Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
4.15 Model fits to Balmer and Oxygen Emission Lines in Latest CL-LINER
Follow-up Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
4.16 XMM-Newton Light Curves of CL-LINERs . . . . . . . . . . . . . . . 159
4.17 NuSTAR Light Curves of CL-LINERs . . . . . . . . . . . . . . . . . . 160
xiv
List of Abbreviations
AGN Active Galactic Nucleus
BH Black Hole
BOSS Baryon Oscillation Spectroscopic Survey
CLAGN Changing-look AGN
CL-LINER Changing-look LINER
DESI Dark Energy Spectroscopic Instrument
DRW Damped Random Walk
FPM Focal Plane Module
HST Hubble Space Telescope
IR Infrared
LIER Low-ionization Emission Line Region
LINER Low-ionization Nuclear Emission Line Region
MIR Mid-infrared
MSIP Mid-Scale Innovations Program
NASA National Aeronautics and Space Administration
NLSy1 Narrow-Line Seyfert 1 Galaxy
NuSTAR Nuclear Spectroscopic Telescope Array
P48 Palomar Oschin 48-inch Schmidt Telescope
P60 Palomar 60-inch SED Machine
SDSS Sloan Digital Sky Survey
SMBH Supermassive Black Hole
SN Supernova
SPIDERS SPectroscopic IDentfication of ERosita Sources
Sy Seyfert galaxy type
TDSS Time-Domain Spectroscopic Survey
TESS Transiting Exoplanet Survey Satellite
TNS Transient Name Server
ToO Target of Opportunity
UV Ultra-violet
XMM X-ray Multi-Mirror Mission
Table 1: Abbreviations Used In This Thesis
For more abbreviations related to software and facilities, see Appendix A.
xv
Chapter 1: Introduction
1.1 Introduction to Supermassive Black Holes (SMBHs) and Active
Galactic Nuclei (AGN)
Black holes (BHs) are mathematically a singularity in space-time, and astro-
physically the runaway gravitational collapse of dense remnants of the explosive end
states of massive stars tens to hundreds of times the mass of the sun. The scaled-up
counterparts to these stellar mass black hole systems are supermassive black holes
(SMBHs), which are typically millions to billions of times Solar. These black holes
have more mysterious origins for growth to these sizes, yet are found to comprise
the centers of nearly all moderate to large sized galaxies ? some with companion
SMBHs, and others largely dormant in the current era / nearby universe (for ex-
ample the BH located at the Milky Way?s galactic center, Sagittarius A*). Stellar
mass black holes actively accreting1 from supergiant stellar companions as persis-
tent sources of high-energy radiation, so-called X-ray binaries, have supermassive
counterparts in the form of active galactic nuclei (AGN), which occupy roughly 10%
of galaxies. The AGN engine is thought to be driven by accretion of a reservoir of
primarily gaseous ionized material gravitationally drawn from the surrounding host
galaxy into a disk around the central SMBH.
1Accretion refers to the accruing of matter through gravitational force leading to the black
hole?s growth.
1
1.1.1 Anatomy of an AGN
The accretion onto the central SMBH taking place in AGN is the most efficient
natural conversion of mass to energy known. AGN are also the most luminous
persistent sources of electromagnetic radiation in the observable universe, spanning
wavebands from gamma-rays to radio. Among the most massive types of AGN
systems (& 108M), quasars (historically shortened from ?quasi-stellar objects?2)
can reach luminosities3 of 1047 erg s?1. These outshine the combined light from all
the tens of billions of stars occupying their host galaxies. Over the lifetime of the
quasar, this output can surpass the gravitational binding energy of the galaxy by
two orders of magnitude (Silk & Rees, 1998).
AGN produce radiation at all wavelengths, representing many diverse ener-
getic processes: from gamma rays due to annihilation of near-relativistic energetic
particles near the event horizon, to UV/optical photons from viscous heating of
the infalling material, to X-rays as a result of magnetically confined plasma, to
synchrotron radiation produced by jets launched as a result of strongly wound mag-
netic field lines coupled to the black hole spin. Section 1.2.1 discusses in detail the
classification scheme for AGN types based on observed radiation. This dissertation
focuses mainly on optically-selected, radio-quiet, UV- and X-ray-bright sources, and
therefore this section focuses on the primary source of that emission ? the com-
plex structures formed by the accretion flow. Information about these components
is partially derived from intensive studies across many wavelengths of stellar mass
black hole systems, which accrete from nearby giant stars rather than a reservoir of
gas in a host galaxy?s center, and therefore provide a solid theoretical basis upon
which to build this work (see e.g. Ruan et al. 2019a, Figure 1.1). However, ways in
which this scaling by mass can break down are discussed further in Section 1.2.
2QSOs, or quasi-stellar objects, are named due to their brightness in visible wavelengths being
historically confounded with stellar sources despite being millions of times more distant.
3Luminosity refers to the intrinsic rate of energy release due to radiation of a source.
2
Figure 5. from The Analogous Structure of Accretion Flows in Supermassive and Stellar Mass Black Holes: New Insights from Faded Changing-
look Quasars
null 2019 APJ 883 76 doi:10.3847/1538-4357/ab3c1a
http://dx.doi.org/10.3847/1538-4357/ab3c1a
? 2019. The American Astronomical Society. All rights reserved.
Figure 1.1: Upper panels: An example of how the accretion behavior of an SMBH
in an AGN system may be described by that of an X-ray binary. Figures from
Ruan et al. (2019a,b) show the predicted ?OX (ratio of UV and X-ray flux densities)
and Eddington ratio (accretion rate parameter) evolution in AGN analogous to an
X-ray binary outburst (upper left). Comparison to changing-look AGN (including a
changing-look LINER presented in Chapter 2) appears to show it following the shape
of the XRB outburst evolution through this parameter space (upper right). Lower
panel: Diagram of accretion onto a black hole. One such interpretation of the Ruan
et al. (2019b) result is based on a scaled model of the spectral evolutionary sequence
through accretion states defined by luminosity and Eddington rate. Components
such as the X-ray emitting corona and the UV emitting disk are described in the
following Section.
3
As particles sink toward the black hole gravitational well following general
relativistically curved trajectories along space-time, they flatten on average due to
momentum conservation into a disk-like structure known as the accretion disk. The
matter approaching the event horizon ? the so-called ?point of no return? ? of the
black hole is high temperature (T ? 104 K), viscous and turbulent, and threaded by
strong magnetic field lines. As such, it is an ideal and sensitive laboratory for physics
in extreme regimes of gravity and temperature far beyond what can be achieved on
Earth.
Observables from AGN can be decomposed into a number of interrelated com-
ponents, relevant at differing energies, centered on a supermassive black hole sur-
rounded by hot plasma and a radiatively-efficient geometrically thin and optically
thick accretion disk emitting primarily in the optical-UV (Malkan, 1983). We review
a subset of relevant components in the following sections.
1.1.1.1 The X-ray Emitting Corona
A tenuous component called the corona emits X-rays from above and below
the innermost accretion flow, and may result from the confluence of plasma flow-
ing along magnetic field lines threading the disk (Galeev et al., 1979; Merloni &
Fabian, 2001). Again, this dissertation focuses on modes of accretion in radio-quiet
systems, although some AGN may display evidence for disk winds in addition to
radio and gamma ray emitting synchrotron powered jets. These are powered by
the Blandford-Znajek mechanism, which magnetically extracts angular momentum
from the spanning black hole and converts its rotational energy to electromagnetic
power (Blandford & Znajek, 1977), and thought in some models to be connected to
the corona as the base of a weak or failed jet (e.g. Wilkins et al. 2015). The exact
geometry of the corona has not yet been unambiguously measured for any system.
In the lamp-post geometry, the X-ray emitting region is approximated as a point
4
source above a thin disk and along the axis of azimuthal symmetry. This model is
not as physical as those accounting for the emissivity structure of a geometrically
extended corona, but has been shown to describe observed accretion disk irradiation
profiles well, so this is what is often assumed. This hot component is required based
on the temperature of the disk (assuming it is emitting as a blackbody) being too
cold to explain observed X-ray emission from its vicinity (Ghisellini, 2013). Ob-
served X-ray energies (? 100 keV) indicate thermal temperatures of the system of
T ? 107 K. The intrinsic X-ray emission from the corona has a spectrum in the
form of a power law.
1.1.1.2 The Soft X-ray Excess
Primarily in high-mass-accretion-rate AGN, a soft-X-ray excess in flux be-
low ?1-2.5 keV is observed to deviate from the power law continuum at varying
strengths, depending somewhat on the level of absorption affecting the observation
(Arnaud et al., 1985; Singh et al., 1985; Turner & Pounds, 1989).
This ?soft excess? can be modeled physically as a reflection component made
up of emission lines from the inner ionized disk which have been relativistically
broadened beyond identification (Crummy et al., 2006; Gierlin?ski & Done, 2004),
an inner Compton upscattering region of the disk powered by the accretion flow
(Done et al., 2012), or an additional power law component due to the launching of a
jet (Chatterjee et al., 2009; Kataoka et al., 2007) or a number of other diverse ideas
that have been proposed; however, its origin remains largely mysterious and best-fit
models are largely degenerate (Lohfink et al. (2012); Page et al. (2004).
1.1.1.3 Gas Clouds and Obscuring Structures in AGN
The AGN systems display spectroscopic evidence for additional cloud-like nu-
clear structures in the infrared (IR), optical and X-rays called the broad line emitting
5
region (BLR)4, the narrow line region (NLR), and the clumpy, dusty molecular torus
encircling the accretion flow, the radius of which is set by the sublimation temper-
ature of the dust. These components are all discussed further in Section 1.2.1. A
subset of AGN displaying broad lines show double-peaked profiles, which can be in-
terpreted in the context of either a Keplerian disk model (e.g. Eracleous & Halpern
2003), or a dual AGN scenario in rare instances that they cannot be explained as
a result of typical disk dynamics (e.g. Ve?ron-Cetty & Ve?ron 2000.) Obscuration
by gas and dust along the line of sight to the accretion flow in the form of out-
flowing winds also affect AGN spectra in complex ways that present a challenge to
disentangle and interpret. For these structures spanning large ranges in mass and
intrinsic physical separation, time resolution is much higher than spatial resolution.
Therefore, time-domain investigations such as those laid out here are crucial for a
complete understanding of AGN phenomena.
1.1.2 AGN Variability
High energy radiation from the innermost regions of AGN can show variability
on timescales down to minutes and even seconds. Decades long observation cam-
paigns of well-studied, relatively nearby AGN have shown that the optical variability
caused by the chaotic accretion process can be modeled well by a damped random
walk (DRW; e.g. Kelly et al. 2009). Most rapid5 variability will occur on the orbital
timescale (a.k.a. the dynamical time scale), where rS is the Schwarzschild radius or
r 2GMBHS = c2
?
2?R R3 MBH R
torb = = 2? = 1.1 days( )( )
3/2 (1.1)
v? GM 108M 100 rS
4This refers to the nuclear region producing broad hydrogen Balmer series emission lines ob-
served in optical spectra of AGN.
5In this case, the term ?rapid? for AGN is relative and refers to phenomena that is observable
on human timescales.
6
6 or the light-crossing time, although propagating fluctuations, accretion disk insta-
bilities, and variable obscuration can also contribute (see discussion in Section 1.2.4
of this Chapter).
Accreting stellar mass black hole systems are observed to display variability in
both brightness and spectral slopes on predictable cycles related to drastic changes
in accretion states lasting tens of days (populating the so-called ?turtle head? di-
agram). Scaling viscous timescales7 simply from such systems, we may expect to
observe more extreme forms of variability from AGN related to changes in accretion
states on the order of thousands to hundreds of thousands of years (Lawrence, 2018;
Siemiginowska et al., 1996).
1.2 Changing-look AGN: A Unique Challenge to AGN Theory
1.2.1 The AGN Unification Scheme and Select Host Galaxy Types
Broad and narrow line regions of AGN are named as such due to the observa-
tions of widths of Hydrogen features observed in spectra of the regions probed by
these components, thought to be clouds of gaseous material. Type 1 and 2 AGN
are classified according to the presence or absence of these lines, thought to be sig-
natures of velocity broadening due to proximity of the clouds to the strong gravity
near the central SMBH. A FWHM of 2000 km s?1 is usually quoted as the boundary
for defining Type 1 vs. Type 2 AGN.
Antonucci (1993) presented a unification scheme (see Figure 1.2) following
observations of type 2 AGN in polarized light which showed their spectra looked
much like that of type 1 AGN. From this evidence, it was established that the
6This timescale is only accurate many gravitational radii (r = GMG c2 ) from the black hole, before
general relativistic effects become important.
7The viscous timescale is the time required to diffuse momentum via thermal collisions of gas
particles in a hot medium, t = 107 seconds ??1(r/10h )2(R/3r )3/2(M 8visc d S BH/10 M), where hd is
the disk thickness.
7
view along the line of sight to the observer was the source of the discrepancy, that
narrow line AGN were obscured by the torus and broad line AGN were being viewed
closer in to the central SMBH (Rowan-Robinson, 1977; Sheng et al., 2017; Urry &
Padovani, 1995). Therefore, for many decades, Seyfert Type 2 galaxies were thought
to be obscured due to being observed at high inclination angles to the torus.
Figure 1.2: The unification model of AGN, adapted from Urry & Padovani (1995),
is based on orientation to the line of sight to the observer as well as by the presence
or absence of components such as a radio-loud jet.9
Under this unification model, the diverse zoo of AGN observational classes:
Quasars, QSOs, Seyfert galaxies, BL Lac objects, blazars, can all be described by
the same unified model: that all AGN are powered by the same mechanism with
differences in the vantage point to the observer of the various physical components
contributing persistent emission across the electromagnetic spectrum comprising
9Figure hosted at https://fermi.gsfc.nasa.gov/science/eteu/agn/figure1.jpg.
8
the accretion flow (e.g. gas, dust) and/or a jet (e.g. plasma, magnetic fields). See
Section 1.1.1 for more details about these components.
In the following sections we focus on introducing the host galaxy types most
relevant for the work laid out in this thesis.
1.2.2 Intermediate AGN Type Classifications
Many AGN spectra do not fall neatly into either a type 1 or type 2 subclass.
Therefore, the presence or absence of broad components in the highest order Balmer
emission lines in the series, H? and H?, defines a number of intermediate type
designations between broad (1) and narrow (2). Osterbrock (1981) specified the
difference between Seyfert 1.5, 1.8 and 1.9 as follows:
? Type 1.9 ? only a broad component in the H? line, narrow higher order
Balmer lines.
? Type 1.8 ? broad component in the H? line, a relatively weak broad compo-
nent in H?.
? Type 1.5 ? strength of broad components are comparable between H? and
H?.
Winkler (1992) expanded upon this scheme to quantitatively distinguish the
subtypes based on the flux ratio of H? to [O III]?5007, with cutoff classes at 0.33,
2.0 and 5.0 (including Type 1.2). At times these intermediate labels have also
classified non-Seyfert objects such as low ionization nuclear emission line region
galaxies (LINERs) based on the appearance of weak emission lines in their stellar
dominated optical spectra (see Section 1.2.4 for a full description of this class).
9
1.2.3 Narrow Line Seyfert 1 Galaxies
Narrow-line Seyfert 1 type galaxies (or NLSy1s), are a separate class of highly
X-ray variable AGN with observed Balmer line widths approximately intermediate
between 1000 and 2000 km s?1 (Goodrich, 1989). They therefore have derived virial
black hole mass (MBH,vir) estimates (Shen et al., 2011) systematically smaller than
that of normal Seyfert type 1 AGN,
( )( )2
? RBLR FWHM(H?)MBH,vir = 1.5 105 M
light days 103kms?1
where RBLR = 32.9(
?L5100A )0.744 ?1 light days and L5100A is the luminosity density at10 ergs
5100 A? (Kaspi et al., 2000).
Although multiwavelength evidence points to NLSy1s accreting close to the
Eddington limit10 (e.g. Xu et al. 2012, Netzer & Trakhtenbrot 2007 and references
therein), the virial black hole mass based on H? measurements from which this
property has partially been derived has been posited to be due to an inclination effect
by Rakshit et al. (2017) (see Section 2.3.3 for further discussion and explanations).
NLSy1s often display relatively prominent Fe II complexes in their optical spectra
(Osterbrock & Pogge, 1985; Rakshit et al., 2017), which, when explained by an
increased covering factor of the BLR (Ve?ron-Cetty & Ve?ron, 2000), is in agreement
with the high accretion rate interpretation along the axes of the eigenvector relation
classification scheme for AGN that also depends upon orientation (e.g. Shen & Ho
2014).
NLSy1s are often more variable in the X-rays than the optical. However, dur-
ing the ASASSN survey, Trakhtenbrot et al. (2019a) reported on a nuclear optical
transient with persistent UV-bright emission in a NLSy1 host galaxy and associated
it with two other similar events in the literature. The multiwavelength properties
10The Eddington limit is the accretion rate at which the radiation force emitted from the ac-
creting body balances the gravitational force on the accreted matter.
10
of this transient did not fit within the framework of previously known transient
phenomena that are unrelated to AGN variability such as a tidal disruption event
(TDE) or supernova (SN), and was instead interpreted as an dramatic enhancement
in accretion rate in a pre-existing AGN. Kankare et al. (2017) had previously re-
ported on a similar population of flaring events occurring within NLSy1 galaxies. In
Chapter 3 we report several more of these events occurring in known and candidate
NLSy1s discovered by ZTF.
1.2.4 The Heterogeneous Population of Low Ionization Nuclear Emis-
sion Line Regions (LINERs)
The optical spectra of star forming galaxies with strong Balmer lines and
low ionization spectral features can often be confused for narrow-line AGN. The
Baldwin, Phillips & Terlevich, or BPT, diagram (Baldwin et al. 1981, Figure 1.3)
helps to distinguish these along with a third galaxy type: LINERs, or low ionization
nuclear emission line regions.
LINERs, as their name implies, display narrow low excitation emission lines,
as well as a weak nonthermal continuum. We show an example LINER host galaxy
SDSS spectrum compared with a quasar spectrum in Figure 1.5. Some UV, X-
ray, and radio studies showing compact and variable emission may indicate a low
luminosity AGN underlying the stellar continuum (e.g. Maoz 2007). They are the
largest local AGN subpopulation, representing nearly 30% of all nearby galaxies
(Heckman, 1980).
A separate class of ?LIERs? have emerged, with low-ionization emission not
from the nucleus of the galaxy but rather from processes in non-nuclear stellar
populations which result in the same empirical line ratios as LINERs (Belfiore et al.,
2016).
The observational subclasses of AGN determined by emission line features dis-
11
Seyfert Galaxies
LINERs
Star-forming
Galaxies
Figure 1.3: Galaxies from the Sloan Digital Sky Survey (SDSS) categorized based on
the Baldwin, Phillips & Terlevich (BPT) emission line ratio diagnostic diagram.12
Additional mappings from Kauffmann et al. (2003) and Kewley et al. (2001) are
shown here in orange. This combined classification scheme serves to distinguish
galaxies displaying primarily stellar emission (star-forming galaxies) from accretion
onto supermassive black holes as the dominant emission mechanism (LINERs and
Seyfert galaxies).
12
cussed in this section divide the population of radio-quiet AGN. AGN are more
broadly divided by their emission characteristics, into the so called ?radiative-
mode,? ?wind-mode,? or ?quasar-mode? for those with higher accretion rates and
luminosities primarily from a disk13, and those dominated by outflows or jets in
the ?radio-mode?. The presence or absence of strong radio emission from an AGN
may be linked to the black hole spin parameter (e.g. Ve?ron-Cetty & Ve?ron 2000
and references therein). This physical property may also play an analogous role in
distinguishing LINERs, and may play a role in hosting environments conducive to
accretion state changes such as in changing-look AGN (Dodd et al. 2021; see also
Section 1.2).
UV and X-ray studies of LINERs indicate that they may be weak i.e. low-
accretion-rate or low-radiative-efficiency analogs of Seyfert AGN, and therefore lower
density accretion disks (e.g. Nemmen et al. 2014), although this notion was chal-
lenged by Maoz (2007) who argued for the pure thin disk scenario based on UV/X-
ray observations of LINER nuclei.
1.2.5 Review of Changing-Look AGN
According to the AGN unification scheme laid out in Section 1.2.1, type 2 AGN
are obscured by a dusty molecular torus, and expected to not display variability. Five
decades ago, Seyfert 2 Mrk 6 displayed optical continuum variations accompanied
by the dramatic appearance of broad Balmer emission lines, changing its optical
taxonomy to a type 1 (Khachikian & Weedman, 1971; Pronik & Chuvaev, 1972). In
subsequent years, several more puzzling contradictions to this framework emerged.
Changing-look AGN14 (CLAGN) demonstrate large-amplitude optical and/or
X-ray variability accompanied by a spectrum that completely changes its heretofore
13See Section 2.2.1 for further discussion on empirical distinctions between quasars and other
AGN such as Seyfert galaxies.
14Optical changing-look AGN are distinct from X-ray AGN that exhibit ?changing-look? tran-
sitions between Compton-thick and -thin classes.
13
dichotomous empirical classification from a broad-line AGN with a blue continuum,
to an active galaxy with narrow emission lines (or vice-versa).
A small number of changing-look AGN had been discovered in the mid 20th
century, however the advent of systematic searches through large scale photometric
and spectroscopic surveys for extremely variable quasars such as PanSTARRs and
TDSS/SDSS greatly improved on this number (MacLeod et al., 2016; Ruan et al.,
2016; Rumbaugh et al., 2018).
The first changing look quasar SDSS J015957.64+003310.5 (redshift z = 0.31)
was reported by (LaMassa et al., 2015). Previously a Type 1 AGN in SDSS, it was
found in a subsequent SDSS BOSS (Dawson et al., 2013) epoch to have transitioned
to a Type 1.9 AGN in a 9 year period. The source was also detected by both
archival Chandra and XMM-Newton observations within the Stripe 82X survey
(LaMassa et al., 2013a,b). They systematically ruled out other explanations for the
dramatic spectroscopic and continuum transition such as obscuration by an orbiting
cloud of gaseous material. They also linked CLAGN to an evolutionary theory of
the unified AGN intermediate types being on an evolutionary track dependent on
accretion rate and black hole mass, which we confirm in Section 2.3.5 for the sample
presented in this dissertation. Elitzur et al. (2014) predict a natural sequence (shown
in Figure 1.4) in which AGN transition from type 1 to 1.2/1.5 to 1.8/1.9 due to
a change in accretion rate and the availability of ionizing radiation affecting the
emission line intensities.
Another of the known CLAGN, iPTF16bco (Gezari et al., 2017) was caught
?turning-on? in the iPTF survey into a broad-line quasar from a LINER galaxy
(Figure 1.5). iPTF16bco was one of only three cases of a CLAGN in a LINER out
of the nearly 70 known CLAGN at the time.Furthermore, as a LINER, iPTF16bco
had a lower inferred accretion rate in its low state (L/LEdd . 0.005, Gezari et al.,
2017) compared to the majority of previously discovered CLAGN (MacLeod et al.,
14
2019), implying a much more dramatic transformation.
Figure 1.4: Posited transition between AGN spectral types through L vs. MBH and
M? parameter spaces, from type 1 (blue, filled) to type 2 (red open circles), driven by
the accretion rate and resulting in this luminosity-emission relation. We trace the
relatively rapid (years long) evolution from LINER to type 1 QSO through this same
parameter space for the CL-LINER sample presented in Chapter 2, some increasing
in luminosity by at least an order of magnitude during the transition. From Elitzur
et al. (2014).
These discoveries of unique objects were followed by a slew of observational
studies across the wavebands to understand what each AGN emission region could
illuminate about the changing-state mechanism, e.g. Parker et al. (2016); Stern
et al. (2018), in tandem with theoretical explorations into how other established
fields, such as stellar mass BHs, could inform interpretation of these AGN events,
(Noda & Done, 2018; Ross et al., 2018). These studies bolstered the idea that
CLAGN were in fact driven by changes in the accretion rate and that they could
occur on human timescales. In the meantime, a growing number of continuing
15
systematic searches proved the commonality and heterogeneity of these systems in
Seyfert AGN and quasars.
[OII] H? H? H? [OIII] H? + [NII]
Keck2 2016 Jun 04
60 SDSS 2004 Jun 16
40
20
0
3000 4000 5000 6000 7000 8000
Rest Wavelength (A?)
Figure 1.5: Comparison of early and follow-up spectra of iPTF 16bco showed the
appearance of broad lines where there were previously none, in concurrence with
a smooth, slowly rising blue light curve with a plateau in the final months of the
intermediate Palomar Transient Facility Survey.
These observations challenge the established theory of the AGN unification
scheme, which attributes these spectral differences solely to the orientation of the
line of sight to an obscuring toroid of gas and dust (Antonucci 1993; See Section 1.1.1
for more details). Typically, the difference between type 1 and type 2 AGN has been
explained in this way by geometry alone, but such changes necessitate a paradigm
shift. However, the mechanism behind these changes are not well understood (e.g.
Stern et al. 2018). Recent studies have shown that these transformations from one
AGN subtype to another in a single object, observed on human timescales, are
more rapid than expected from accretion theory by a factor of ten thousand (e.g.
Lawrence 2018). Evidence suggests these events may be the results of environmental
changes due to instabilities propagating through the accretion flow (e.g. Ross et al.
2018; Yan et al. 2019; see Figure 1.6), but the triggering mechanism driving these
sudden flares is not well understood.
The nature of the CLAGN spectral transformation is most often attributed to
16
F? (10
?17 erg/cm2/s/A?)
Figure 1.6: Model from Figure 4 in Ross et al. (2018) illustrating possible accre-
tion modes throughout several years of follow-up observations for changing-look
quasar SDSS J110057.70-005304.5. The physical mechanism for triggering propa-
gating thermal fronts throughout the disk is not well understood, however, such an
unstable configuration may explain the years- to decades-long timescales for transi-
tions between spectral classifications observed in changing-look AGN.
17
changes in accretion rate (e.g. Elitzur et al. 2014, LaMassa et al. 2015, MacLeod
et al. 2016, Gezari et al. 2017, Rumbaugh et al. 2018). Shappee et al. (2014) proposed
that observed X-ray through NIR lags in CLAGN were the result of reprocessing of
variable X-ray illumination of the accretion disk, and Sheng et al. (2017) reported
mid infrared variability lagging that of the optical in support of dust reprocessing.
Tidal disruption events, which occur when a star is tidally destroyed by the
gravitational influence of a black hole, are another way a dormant SMBH might
feed sporadically on stars, although TDEs likely also occur in variable AGN hosts
Jiang et al. (2019); Merloni et al. (2015). Although ZTF has vastly improved our
knowledge of this accretion event type in the last 3 years (see van Velzen et al. 2020b
and references therein) many hurdles to observing TDEs in the foreground of AGN
hosts remain, and they are therefore often found serendipitously (see Sections 1 and
4.4.3 of Chapter 3 for further discussion.) However, MacLeod et al. (2016); Ruan
et al. (2016) and others showed that TDEs could not fully explain CLAGN, in part
due to the existence of CLAGN growing fainter rather than brighter in nearly equal
(albeit small) numbers.
Still, theoretical challenges abound for all proposed scenarios, and further
observations are needed to examine the full range of multi-wavelength phenomena
occurring during these accretion state changes and within their hosts.
1.3 Overview of Thesis Work
1.3.1 Why Study AGN with ZTF?
Astronomers more interested in transients fondly refer to AGN as the ?extra-
galactic vermin? of the sky, and often much effort goes into rejecting them from sur-
veys as uninteresting sources that nevertheless persistently ?go bump in the night?.
However, difference imaging is well situated to pick out significant variability above
18
the low level baseline stochasticity of most broad line AGN, or to pick out variability
from AGN which are not typically extremely variable in the optical (e.g. NLSy1s
and other narrow-line AGN). ZTF autonomously provides densely sampled epochs,
and therefore is able to identify individual features in light curves. The vast amount
of archival optical spectroscopy made available by the Sloan Digital Sky Survey
(SDSS) coupled with the Spectral Energy Distribution Machine (SEDM) robotic
spectrograph?s capabilities to pick out broad lines in follow-up spectra makes the
Zwicky Transient Facility the perfect search engine for discovering CLAGN sources
in real time. For example, 16bco was discovered serendipitously in the final months
of a search for tidal disruption events (TDEs) in the intermediate Palomar Tran-
sient Factory (iPTF) survey to be a quasar turning on from a quiescent, early-type,
LINER-like galaxy (Figure 1.5; Gezari et al. 2017), along with a similar source dis-
covered in archival data by Yan et al. (2019). They predicted several more of these
sources would be discoverable in the order-of-magnitude increase in survey volume
achievable by ZTF. Current detection methods are also well-suited for finding this
particular class of CLAGN turning on, due to the high percentage of quiescent types
comprising galaxies in the nearby Universe? (Dodd et al., 2021).
1.3.2 Summary of Chapters
This dissertation consists of multiwavelength real-time investigations of ex-
treme transients associated with sudden changes in the environments of the SMBHs
in galaxy centers. Possible modes of SMBH accretion we ascribe to these outbursts
span from the tidal disruption of stars in AGN, to the curious established and grow-
ing class of CLAGN. CLAGN demonstrate large-amplitude optical and/or X-ray
variability accompanied by a spectrum that completely changes its heretofore di-
chotomous empirical classification from a broad-line AGN with a blue continuum,
to an active galaxy with narrow emission lines (or vice-versa), challenging the AGN
19
unification scheme based on the orientation of the line of sight to an obscuring
torus. These transformations from one AGN subtype to another, observed on hu-
man timescales, are more rapid than expected from accretion theory by a factor
of ten thousand. Evidence suggests these events are the results of accretion state
changes due to instabilities in a pre-existing disk, but the mechanism driving these
sudden changes is not well understood.
With the ZTF Northern Sky Survey, coupled with data from various spectro-
scopic and photometric follow-up resources (primarily the Lowell Discovery Tele-
scope), we sought to understand the nature of the AGN undergoing flares and
the appearance of spectral features. Informed by an archival iPTF pilot study
of CLAGN, and with ZTF?s upgraded suite of automated telescopes monitoring the
entire Northern sky and streaming millions of data points from new events on a
nightly cadence, we have been able to study an unprecedented volume of dramatic
activity from galaxy centers in real time.
This work has made two main contributions to the field of AGN time-domain
science, with discoveries resulting in a paradigm shift about the timescales and host
galaxies of extreme accretion state changes:
1. Establishing a new observational class of quiescent galaxies caught turning
into quasars, and
2. Linking a (separate) new observational class of unusually smooth and rapid
outbursts to a subclass of active host galaxies undergoing enhanced accretion,
while more than tripling the number of such known events in the literature.
Both transient classes emerged from ongoing searches for CLAGN and TDE phenom-
ena, and are associated with subtypes of AGN with interesting, and still not fully
understood, properties. We found that the former class of objects display spectro-
scopic transformations more dramatic by an order of magnitude when hosted in low
20
ionization nuclear emission line region (LINER) galaxies, and the latter occur pref-
erentially in Narrow-Line Seyfert 1 (NLSy1) galaxies. Chapter 2 of this dissertation
was published in a first-author Astrophysical Journal article entitled ?A New Class
of Changing-look LINERs? (Frederick et al., 2019). Chapter 3 of this dissertation,
(Frederick et al., 2020) entitled ?A Family Tree of Optical Transients from Narrow-
Line Seyfert 1 Galaxies,? links novel findings to other NLSy1-associated transients
in the literature, and provides a framework for spectroscopically classifying such
flares in future wide-field surveys.
The fourth chapter of this dissertation, entitled ?The X-ray View of a New
Class of Changing-look LINERs? (Frederick et al. 2021, in prep.), includes the 3-
year catalog of this new class of objects discovered with ZTF, never-before-seen at
high X-ray energies, and describes the results of a year-long joint XMM-Newton
and NuSTAR observing program entitled ?The First X-ray View of a New Class of
Changing-Look AGN with XMM and NuSTAR? (PI: Frederick). To complement the
hard X-ray data, we also collected target-of-opportunity UV and X-ray observations
of this growing sample with the Neil Gehrels Swift Observatory, as well as multi-
epoch optical spectroscopic follow-up with the Gemini GMOS-N Observatory, as
part of a proposal entitled ?The Real-Time Appearance of the BLR in a New Class
of Changing-Look LINERs Discovered by ZTF? (PI: Frederick), for which multi-
epoch optical spectra were collected in early 2020. This ongoing study will test
physical mechanisms driving this new class of changing-look LINERs by mapping
the structure of the accretion flow state changes.
In summary, in Chapter 2, we establish a new class of changing-look LIN-
ERs and present follow-up observations for a candidate state transition that was
discovered to be occurring in real-time. In Chapter 3, we link a number of optical
transients in AGN to NLSy1 host galaxies, including a changing-look LINER from
the sample presented in Chapter 2, and provide a framework for classification in
21
future surveys. In Chapter 4, we present preliminary analysis of X-ray follow-up
observations of the full CL-LINER sample. Chapter 5 summarizes the findings of
this dissertation and discusses next steps for this work.
ZTF, using difference imaging as a discovery mechanism, has allowed for real-
time detection and rapid multiwavelength follow-up of events related to accretion
onto supermassive black holes like never before. In this dissertation, we establish
that certain nuclear environments are more conducive to hosting these newly discov-
ered dramatic transients, not only to better understand the accretion phases of their
evolution, but also to improve the efficiency of upcoming generations of real-time
searches for the most dramatic AGN outbursts with large-area facilities such as the
Vera C. Rubin Observatory and the fifth Sloan Digital Sky Survey.
22
Chapter 2: A New Class of Changing-Look LINERs
We report the discovery of six active galactic nuclei (AGN) caught ?turning
on? during the first nine months of the Zwicky Transient Facility (ZTF) survey.
The host galaxies were classified as LINERs by weak narrow forbidden line emission
in their archival SDSS spectra, and detected by ZTF as nuclear transients. In five
of the cases, we found via follow-up spectroscopy that they had transformed into
broad-line AGN, reminiscent of the changing-look LINER iPTF16bco. In one case,
ZTF18aajupnt/AT2018dyk, follow-up HST UV and ground-based optical spectra
revealed the transformation into a narrow-line Seyfert 1 (NLS1) with strong [Fe VII,
X, XIV] and He II ?4686 coronal lines. Swift monitoring observations of this source
reveal bright UV emission that tracks the optical flare, accompanied by a luminous
soft X-ray flare that peaks ?60 days later. Spitzer follow-up observations also detect
a luminous mid-infrared flare implying a large covering fraction of dust. Archival
light curves of the entire sample from CRTS, ATLAS, and ASAS-SN constrain the
onset of the optical nuclear flaring from a prolonged quiescent state. Here we present
the systematic selection and follow-up of this new class of changing-look LINERs,
compare their properties to previously reported changing-look Seyfert galaxies, and
conclude that they are a unique class of transients related to physical processes
associated with the LINER accretion state.
23
2.1 Introduction
The observed diversity in the optical spectra of AGN, with well-defined sys-
tematic trends known as the eigenvector relations, are understood to be a function of
both orientation as well as accretion rate (e.g. Shen & Ho 2014). ?Changing-look?
active galactic nuclei (CLAGN) are a growing class of objects that are a challenge
to this simple picture, in that they demonstrate the appearance (or disappearance)
of broad emission lines and a non-stellar continuum, changing their classification
between type 1.8-2 (narrow-line) to type 1 (broad-line) AGN (or vice versa) on a
timescale of years. The nature of this spectral transformation is most often at-
tributed to changes in accretion rate (MacLeod et al., 2016; Oknyansky et al., 2016;
Ruan et al., 2016; Runnoe et al., 2016; Shappee et al., 2014; Sheng et al., 2017),
but the mechanism(s) driving these sudden changes is still not well understood (e.g.
Lawrence 2018; Stern et al. 2018).
One of the known changing-look quasars (CLQs), iPTF16bco (Gezari et al.,
2017), was caught ?turning-on? in the iPTF survey into a broad-line quasar from
a low-ionization nuclear emission-line region galaxy (LINER). LINERs are distin-
guished from Seyfert 2 (Sy 2) spectra via the relatively strong presence of low-
ionization or neutral line emission from [O I] ?6300, [O II] ?3727, [N II] ??6548,
6583, and [S II] ??6717, 6731; a lower [O III] ?5007/H? flux ratio; and a lower
nuclear luminosity. However, the status of LINERs as low-luminosity AGN rem-
nants is a topic of debate, as weak emission in some LINERs could also be powered
by shocks, winds, outflows, or photoionization from post-AGB stellar populations
(Bremer et al., 2013; Filippenko, 1996; Ho et al., 1993; Singh et al., 2013). LINER
galaxies are the largest AGN sub population, and may constitute one-third of all
nearby galaxies (Heckman, 1980; Ho et al., 1997b), yet iPTF16bco was one of only
three cases of a CLAGN in a LINER out of the nearly 70 known CLAGN at the
24
time.1 Furthermore, as a LINER, iPTF16bco had a lower inferred accretion rate
in its low state (L/LEdd . 0.005, Gezari et al., 2017) compared to the majority
of previously discovered CLAGN (MacLeod et al., 2019), implying a much more
dramatic transformation.
We report the discovery of six new CLAGN, all classified as LINER galaxies by
their archival SDSS spectra, detected as nuclear transients by the Zwicky Transient
Facility (ZTF; Bellm et al. (2019a); Graham et al. (2019)), and spectroscopically
confirmed as ?changing-look? to a NLS1 or broad-line (type 1) AGN spectral class.
One of these nuclear transients, ZTF18aajupnt, was initially classified as a candidate
tidal disruption event (TDE) from the presence of Balmer and He II emission lines
(Arcavi et al., 2018). Here, we show that the ZTF light curve, together with our
sequence of follow-up optical spectra and UV and X-ray monitoring with Swift and
follow-up UV spectra with HST, are more consistent with a CLAGN classification. It
was previously thought that, although they are commonly found in Seyferts, coronal
emission lines (such as [Fe VII] ?6088) should never be exhibited by LINER-like
galaxies by definition (e.g. Corbett et al. 1996). However, here we also report the
surprising appearance of coronal lines coincident with an increase in UV/optical and
soft X-ray continuum emission and broad Balmer emission consistent with a NLS1
in this galaxy previously classified as a LINER.
This paper is organized as follows. In Section 2.2, we present our sample
selection of nuclear transients in LINERs, information on the host galaxies, ZTF and
archival optical light curves, optical spectroscopic observations, and multiwavelength
follow-up observations of ZTF18aajupnt, including details of the data reduction
involved. In Section 2.3, we introduce a new class of changing-look LINERs, and
compare their properties to previously reported Seyfert CLAGN, focusing on the
1We note that the other two known so-called CL LINERs, NGC 1097 (Storchi-Bergmann et al.,
1993) and NGC 3065 (Eracleous & Halpern, 2001) are, or are reminiscent of, transient double
peaked emitters, which may be distinct from changing-look AGN.
25
particularly interesting case of ZTF18aajupnt, which transformed from a LINER to
a NLS1. In Section 2.4 we discuss the results of our analysis, the conclusions of
which are summarized in Section 2.5.
Throughout the paper we use UT dates, and assume the following cosmology
for luminosity calculations: H0 = 70 km s
?1 Mpc?1, ??= 0.73 and ?M = 0.27. We
have corrected for Galactic extinction toward the sources where explicitly stated.
All magnitudes are in the AB system, and all uncertainties are at the 1? level unless
otherwise noted. We adopt the definition for a quasar from the SDSS DR7 quasar
catalog (Schneider et al., 2010), as having an apparent i-band PSF magnitude fainter
than 15 and an absolute i-band magnitude brighter than ?22.
2.2 Discovery and Observations
2.2.1 Sample Selection Criteria
We selected CLAGN candidates first flagged as nuclear transients in the ZTF
alert stream (described further in Section 2.2.2) and with a cross-match within
1?.?0 of a LINER or type 2 Seyfert galaxy in the Portsmouth Catalog?s narrow-line
ratio BPT classifications2 (Bolzonella et al., 2000; Thomas et al., 2013). Those
classifications, described further in Section 2.3.1, are based on stellar population
and emission line fits to SDSS DR12 spectra, performed with Penalized Pixel Fitting
(pPXF; Cappellari (2016); Cappellari & Emsellem (2004)) and Gas and Absorption
Line Fitting (GANDALF; Sarzi et al. (2017)), respectively. In this study, we focus
on the ?LINER CLAGN? that emerged as a new class of changing-look AGN and
display the most dramatic spectral variability of the CLAGN in our ZTF sample (we
reserve discussion of the complementary sample of Seyfert CLAGN for a forthcoming
publication).
2https://www.sdss.org/dr12/spectro/galaxy_portsmouth/#kinematics
26
2.2.2 ZTF Light Curve
ZTF surveys the extragalactic3 Northern Sky in two modes: a public Mid-Scale
Innovations Program (MSIP) survey of 15,000 deg2 of sky every 3 nights in g and r
filters, and a high-cadence ZTF partnership survey of 3400 deg2 with a dense cadence
of 6 epochs each in g and r filters per night. It also surveys in i-band every 4 nights
with a footprint of 10725 deg2 (Bellm et al., 2019b). PTF and iPTF (2009?2016;
Law et al. (2009); Rau et al. (2009)) also utilized Palomar Observatory?s Samuel
Oschin 48? Schmidt telescope; the camera upgrade for ZTF has a 47 deg2 FoV and
reaches 20.5 r-band mag in 30 seconds exposures, with a more efficient areal survey
speed of 3760 deg2 hr?1. Images are processed each night by the Infrared Processing
and Analysis Center (IPAC) pipeline (Masci et al., 2019), where difference imaging
and source detection are performed to produce a transient alert stream (Patterson
et al., 2019), distributed to the GROWTH Marshal (Kasliwal et al., 2019) and other
brokers via the University of Washington Kafka system. van Velzen et al. (2019)
presented details of the nuclear transients filtering procedure.
All transients in the sample were discovered in 2018 between April and Novem-
ber, all in the ZTF MSIP survey (specific dates are summarized in Table 2.1).
ZTF18aajupnt4 was also detected in the ZTF Partnership survey on 2018 May 31,
and (as it was detected in both surveys in the same night) was registered publicly
to the Transient Name Server (TNS) as AT2018dyk. Transients were required to
have a real-bogus (RB) score ? 0.5 as classified by ZTF machine learning (Maha-
bal et al., 2019). Further details on the transients, including discovery difference
absolute magnitudes, are in Table 2.1.
3Additional public and private allocations are made to survey the Galactic Plane at higher
cadence. See Bellm et al. (2019b) for details.
4 As ZTF given names are typically a mouthful of letters (appropriately so, due to the require-
ment of naming upwards of a million alerts per night), the ZTF Black Holes Working Group has
informally begun naming TDEs from a fictional world with no shortage of characters: HBO?s Game
of Thrones. As it was initially thought to be a TDE, ZTF18aajupnt was affectionately dubbed
?Tyrion Lannister?.
27
Table 2.1: Basic data for the changing-look LINER sample. We list redshifts from
the Portsmouth SDSS DR12 catalog (Thomas et al., 2013), which is described in
Section 2.2.1. Transition timescales ?t are roughly constrained based on the time
delay between the onset of variability detected in the host in the archival light curves,
and the time of the first spectrum taken in the type 1 AGN state. Estimates of star
formation rate by Chang et al. (2015) are from SDSS+WISE SED model fitting.
?m is the variability magnitude change defined in Eq. 3 of Hung et al. (2018) as
?m = ?2.5log(10?mr,host/2.5 +10?mr/2.5)?mr,host, where mr represents the brightest,
transient r-band magnitude in the difference-imaging light curve. ZTF18aajupnt,
described further in Section 2.3.6 is the least luminous transient, and has the nearest
host of the sample.
Name RA Dec z DLum Discovery Date M
5
Discovery ?t Host Type log SFR ?mvar High State
(hh:mm:ss.ss) (dd:mm:ss.ss) (Mpc) (mag) (yr) [M yr
?1] (mag)
ZTF18aajupnt6 15:33:08.01 +44:32:08.2 0.0367 158 2018 May 317 ?16.59 <0.3 Spiral (SBb D) 0.177 ?0.18 NLS1
ZTF18aasuray8 11:33:55.83 +67:01:08.0 0.0397 171 2018 May 10 ?17.80 <6.8 Spiral (SBa(r)9) 0.147 ?0.06 Seyfert 1
ZTF18aahiqfi10 12:54:03.80 +49:14:52.9 0.0670 296 2018 April 8 ?18.25 <0.6 Elliptical ?0.058 ?0.12 quasar
ZTF18aaidlyq11 09:15:31.06 +48:14:08.0 0.1005 457 2018 April 11 ?19.09 <0.7 Spiral (Sb D) 0.092 ?0.29 quasar
ZTF18aaabltn12 08:17:26.42 +10:12:10.1 0.0458 199 2018 Sept 15 ?17.62 <2.6 Elliptical 0.227 ?0.81 quasar
ZTF18aasszwr13 12:25:50.31 +51:08:46.5 0.1680 813 2018 Nov 1 ?20.40 <5.3 Elliptical 1.267 ?0.72 quasar
The optical photometry for ZTF18aajupnt, ZTF18aasuray, ZTF18aahiqfi, ZTF18aaidlyq,
ZTF18aasszwr, and ZTF18aaabltn is comprised of 398, 200, 35, 35, 143, and 207 im-
ages, respectively, shown in Figure 2.1. We consider only observations with difference
image detections classified as real (with RB score? 0.5 on a scale where 0 is bogus
and 1 is real). The ZTF optical difference imaging light curves show only the tran-
sient nuclear emission in the g- and r-bands. The transients are localized to within
?? +0?.?28 ?? ? ?? ?? +0?.?0.19?0??19 (ZTF18aajupnt), 0.09 0.26 (ZTF18aasuray), 0.11 33?0??11 (ZTF18aahiqfi),. .
0?.?06+0
?.?33 ??
?0??06 (ZTF18aaidlyq), 0
?.?10+0. 20?0??10 (ZTF18aasszwr), and 0
?.?15?0?.?15 (ZTF18aaabltn)
. .
of their host galaxy nuclei, well within our nuclear selection criterion of < 0?.?5.
To quantify the amplitude of the flux increase relative to the host galaxy flux,
and to compare to variability of CLAGN measured from imaging surveys that do
not perform image subtraction, as in Hung et al. (2018), we add the flux of the host
galaxy to the transient flux, to get a variability amplitude, ?mvar = mr,tot?mr,host,
where m = ?2.5 log(10?mr/2.5+?mr,host/2.5r,tot ), and mr represents the brightest tran-
sient ZTF r-band magnitude and mr,host is the archival host magnitude from SDSS
DR14. We find ?mvar values ranging from ?0.12 to ?0.81 mag for the sources
in our sample, with 3 out of 5 below the CLAGN candidate selection criteria of an
28
18 19.25
19.50
19
19.75
20.00
20
r g UVW2 20.25
0 100 200 300 400 0 100 200 300 400
MJD - 2458269 MJD - 2458216
(a) ZTF18aajupnt (b) ZTF18aahiqfi
19.0 17.5
18.0
19.5
18.5
20.0 19.0
19.5
20.5
20.0
0 100 200 300 400 0 100 200 300 400
MJD - 2458218 MJD - 2458248
(c) ZTF18aaidlyq (d) ZTF18aasuray
18.5 17.5
18.0
19.0
18.5
19.5
19.0
20.0 19.5
20.5 20.0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
MJD - 2458348 MJD - 2458263
(e) ZTF18aasszwr (f) ZTF18aaabltn
Figure 2.1: Light curves of the CL LINER sample. Red points represent r-band
difference imaging photometry data taken with the Palomar 48-inch (P48), green
points g-band difference imaging photometry, and the blue points are the UVW2
Swift photometry in the light curve of ZTF18aajupnt, which tracks the plateau in
the optical uncharacteristic of either TDEs or SNe. Note the differences in scale.
29
Apparent Magnitude Apparent Magnitude Apparent Magnitude
Apparent Magnitude Apparent Magnitude Apparent Magnitude
amplitude of ?r > 0.5 mag between SDSS and Pan-STARRS1 imaging observations
adopted by MacLeod et al. (2019).
ZTF18aajupnt (AT2018dyk; discussed more in Section 2.2.5), ZTF18aasuray,
and ZTF18aasszwr display a slow months-long rise and plateau (although a visibil-
ity gap makes this unclear for ZTF18aasuray) with a constant color, and gradual
decline, with ZTF18aasszwr exhibiting a second rise and ZTF18aajupnt growing
redder in the latest observations. All other transients in the sample show flaring
in the light curves (see Figure 2.1) but with less distinct trends, characteristic of
broad-line AGN variability viewed in difference imaging (Choi et al., 2014).
2.2.3 Capturing the Transition in Archival Light Curves
Although difference imaging is a useful real-time discovery mechanism for these
nuclear transients, archival optical photometric observations can fill in the details
of the timing of the transition to its ?on? state. With archival light curves ex-
tending over a baseline of 13 years from the Catalina Real-time Transient Survey
(CRTS; Drake et al., 2009), the All-Sky Automated Survey for Supernovae (ASAS-
SN; Kochanek et al., 2017; Shappee et al., 2014)14, and Asteroid Terrestrial-impact
Last Alert System (ATLAS; Tonry et al., 2018), and ZTF aperture photometry
from the IPAC pipeline measured from the static images, we uncover an intriguing
uniformity in the events (Figure 2.2). Each source in the sample went from lack-
ing any significant variability to flaring dramatically and, for those observed long
enough, subsequently declining (ZTF18aaabltn continues to rise smoothly). As not
all sources in the sample have peaked, we define the transition timescale for each
source reported in Table 2.1 as being from the onset of each flare to the spectro-
scopic confirmation of the appearance of a blue continuum and broad line emission
(except iPTF16bco, for which the onset time was constrained by archival and follow
14http://www.astronomy.ohio-state.edu/~assassin
30
CRTS (|) & ASASSN ( ) ASASSN ( ), ATLAS (+), & ZTF (|)
ZTF18aajupnt + 0.4 mag
14.5 14.5
15.0 ZTF18aasuray + 0.4 mag 15.0
15.5 15.5
ZTF18aahiqfi
16.0 ZTF18aaabltn + 0.3 mag 16.0
16.5 16.5
ZTF18aaidlyq
17.0 17.0
17.5 17.5
ZTF18aasszwr + 0.3 mag Transition from off to on
18.0 18.0
0 2000 4000 0 200 400 600 800 1000 1200 1400 1600
Days since 53466 MJD Days since 57000 MJD
Figure 2.2: Archival light curves of the CL LINER sample summarized in Sec-
tion 2.2.3. The left panel shows years to decades of quiescence (in the ?off?
state while these were still LINER galaxies) observed by CRTS, followed by slow
flares in the faintest sources ZTF18aasszwr, ZTF18aaabltn, and ZTF18aahiqfi.
The right panel shows the rise, flaring, and decline of the sources caught by
ZTF+ATLAS+ASASSN g-band observations at these various stages. The estimated
transition time listed in Section 2.2.3 for each object is marked by a black ???.
When two filters are shown for the same instrument, the redder is shown as more
transparent, as in the case of the ASAS-SN g and V photometric points shown.
up X-ray observations; Gezari et al. (2017)). Turn-on timescales, absolute r-band
magnitudes at the time of detection with ZTF, variability amplitude relative to the
host galaxy flux, and new AGN class following the change are summarized in Ta-
ble 2.1 for all transients in the sample. We discuss the details of each source?s flaring
individually below.
ZTF18aajupnt ? ZTF-matched aperture photometry in g band shows that
ZTF18aajupnt began flaring some time before 2018 March (58200 MJD) ?2 months
prior to discovery in difference imaging on 2018 May 31, and 3 months prior to con-
firmation of a spectroscopic change. The most recent difference imaging photometry
31
PSF Magnitude + Constant
shows a slow decline at constant color. Transition timescale: < 0.3 years, the fastest
in the sample.
ZTF18aahiqfi ? The rise (seen in ZTF g-band matched photometry) starts ap-
proximately at 2017 Sept (58000 MJD), 7 months prior to its spectroscopic change.
It peaks around 2018 May (58250 MJD; ?1 month after discovery with ZTF dif-
ference imaging on 2018 April 8) and subsequently shows a sharp decline. Prior to
this flaring, Catalina Real-time Transient Survey (CRTS shown in the left panel of
Figure 2.2; Drake et al. (2009)) observations in V -band and ASAS-SN showed no
variability above the 0.1 mag level. Transition timescale: < 0.6 years.
ZTF18aaidlyq ? This source displayed a slight flare in ASAS-SN data just
after 2017 Sept (58000 MJD), 7 months prior to detection in ZTF difference imaging
and 8 months prior to spectroscopic confirmation of the existence of a BLR, but was
faint and quiescent in CRTS beginning in 2005 May (note that this source is near a
bright star). Transition timescale: < 0.7 years.
ZTF18aasuray ? Discovery with ZTF difference imaging occurred on 2018
May 10 and shows a slow symmetric rise and decline lasting 300 days. ZTF18aasuray dis-
played flaring in ASAS-SN data beginning around 2011 Aug (55800 MJD), 6.8 years
prior to spectroscopic confirmation of the changing look which occurred on 2018
June 21. Prior to this flaring, Catalina Real-time Transient Survey (CRTS shown
in the left panel of Figure 2.2; Drake et al. (2009)) observations in V -band showed
no variability above the 0.1 mag level. Transition timescale: < 6.8 years.
ZTF18aasszwr ? The rise is visible in CRTS around 2018 July (56500 MJD),
after which it may have plateaued for a time. Most recently there has been a sharp
rise and decline around 2018 May (58250 MJD), with the peak reaching > 1 mag
above original levels. The transition from quiescence thus happened roughly in real
time, and was observed with difference imaging 4 months after the flaring began,
with the spectroscopic change confirmed within 5.3 years of the initial rise time, and
32
within 5 months of the onset of the most recent flare. We note that two decades ago,
ZTF18aasszwr was a variable (rms = 0.14 mJy) radio source between the NRAO
VLA Sky Survey and Faint Images of the Radio Sky at Twenty centimeters (NVSS
and FIRST; Ofek & Frail 2011), with a peak flux density at 1.4 GHz of F? = 2.17
mJy. Transition timescale: < 5.3 years.
ZTF18aaabltn ? CRTS, ATLAS and ASAS-SN show a continuous rise starting
around 2016 April (57500 MJD) but this disregards some slight flaring (by 0.2 mag)
events at 2008 Nov and just before 2014 Dec (57000 MJD), with both returning to
very flat pre-activity levels. This constrains the spectroscopic change to happening
within 1000 days (< 2.7 years) of the flare start time, the first large flare occurring
within 9 months of being observed to be a LINER in 2007 Feb. Transition timescale:
< 2.6 years.
iPTF16bco ? CRTS photometry shows a flare beginning around 2012 March
(56000 MJD), 8 years after being observed to be a LINER and 4 years prior to dis-
covery and classification of a quasar in iPTF, and the latest ZTF g-band data show
it declining rapidly. However, archival XMM Slew Survey observations constrain
the onset of the X-ray source detected by Swift in its broad-line state to < 1.1 years
before (Gezari et al., 2017). Transition timescale: < 1.1 years.
2.2.4 Host Galaxy Morphology
Images of the six transients? host galaxies from SDSS are shown in Figure 2.3,
and basic data including the hosts? names, matched coordinates, redshifts, luminos-
ity distances, morphological types, and star formation rates (SFRs) are summarized
in Table 2.1. The SFR estimates by Chang et al. (2015) were obtained through
Multi-wavelength Analysis of Galaxy Physical Properties (MAGPHYS; da Cunha
et al. (2012)) model fitting of dust extinction/emission, and SEDs constructed from
WISE+SDSS (WISE: Wright et al. (2010)) matched photometry of present-epoch
33
galaxies (we note that SFRs for only two AGN in our sample were measured by
Chang et al. (2015), the rest did not fit their criteria). The bulges of the LINERs?
hosts are similar in apparent color and extent, but the host of ZTF18aaidlyq exhibits
evidence for a bar and ring, and the host of ZTF18aaabltn exhibits apparent elonga-
tion. The host of ZTF18aajupnt stands out in the sample as the only gas-rich spiral
galaxy, and we note that NLS1s typically occur in spiral-type galaxies (Crenshaw
et al., 2003). Black hole masses estimated from the host galaxy luminosity, bulge
mass, and velocity dispersions derived from the SDSS host imaging and spectra have
been measured in Section 2.3.2 and are summarized in Table 2.2.
(a) ZTF18aajupnt (b) ZTF18aahiqfi (c) ZTF18aaidlyq (d) ZTF18aasuray
(e) ZTF18aasszwr (f) ZTF18aaabltn
Figure 2.3: Composite ugriz color SDSS images of the host galaxies of the changing-
look LINER sample. Their individual morphological classifications are listed in
Table 2.1.
2.2.5 Optical Spectroscopy
We obtained spectral follow-up of nuclear transients in known LINERS and
Sy 2 galaxies as described in Section 2.2.1 to confirm changing-look AGN candidates,
as neither ?true? narrow-line Sy 2s nor LINERs are expected to vary significantly.
34
Table 2.2: Properties of the host galaxies of our sample of changing-look LINERs
from ZTF and iPTF. We also show MBH calculated in Section 2.3.2 from the host
galaxy luminosity, mass, and velocity dispersion, respectively.
Name M 15r,host log M
16 17 18 19
Bulge ?F ?L5100A FWHMH log MBH,Mr log MBH,Bulge log M
20 21
? BH,?F
log MBH,vir L/LEdd
(mag) [M] (km s
?1) (1043 erg s?1) (km s?1) [M] [M] [M] [M]
ZTF18aajupnt ?22.00 10.66?0.15 150 0.49?0.11 939?28 8.0 7.8 7.6 6.4 0.09
ZTF18aasuray ?21.70 10.73?0.15 230 10.6?0.4 4270?218 7.9 7.9 8.4 8.0 0.03
ZTF18aaidlyq ?21.64 - 120 12.7?2.3 7726?458 7.9 - 8.2 8.5 0.06
ZTF18aahiqfi ?21.63 - 210 4.1?0.5 8809?723 7.9 - 7.2 8.3 0.2
ZTF18aasszwr ?22.19 11.19?0.15 180 57.0?1.9 6461?846 8.1 8.3 7.9 8.8 0.5
ZTF18aaabltn ?20.62 - 140 0.8?0.2 5195?648 7.3 - 7.5 8.0 0.2
iPTF16bco ?22.21 - 176 17.3?11.0 4183?213 8.4 - 7.9 8.1 0.05
We observed ZTF18aahiqfi, ZTF18aaidlyq, and ZTF18aasuray with the Deveny
spectrograph on the Discovery Channel Telescope (DCT; spectral coverage of 3600-
8000 A?) with a 1?.?5 wide slit, central wavelength of 5800 A? and exposure times
of 2 ? 900, 2 ? 1200, and 1400 seconds on 2018 April 11, May 06, and June 21,
respectively. The DCT spectra were reduced with standard IRAF routines, corrected
for bias and flat-fielding, and combined into a single 2D science frame. Wavelength
and flux calibration were done via a comparison with spectra of an arc lamp and the
flux standard Feige 34, respectively. The spectra have not been corrected for telluric
absorption. We found that the Balmer lines of ZTF18aahiqfi, ZTF18aaidlyq, and
ZTF18aasuray had gotten dramatically stronger and broader compared to archival
SDSS spectra of their hosts, obtained more than a decade prior (in April 2003, Dec
2002, and Feb 2001, respectively).
ZTF18aasszwr and ZTF18aaabltn showed similar striking spectral changes
when they were followed up on 2018 Dec 3 and 9 using the Spectral Energy Distri-
bution Machine (SEDM; Blagorodnova et al. 2018) IFU spectrograph on the Palo-
mar 60-inch (P60; Cenko et al. 2006) operating as part of ZTF. Both displayed
broader emission lines and bluer continuua compared to archival LINER spectra
(from Feb 2007 and Jun 2004, respectively). The SEDM data were reduced with
pySEDM (Rigault et al., 2019).
See the spectral comparisons for all CLAGN in the sample in Figure 2.4, and
zoom-ins of the emission lines in the ?off? states in Figures 2.18, 2.19, and ?on?
35
states in Figure 2.20 of the Appendix. The hosts of all six transients in this sample
were originally classified as LINERs in SDSS, however we re-measured the diagnostic
narrow-line ratios in Section 2.3.1, and find that the majority of the sample is on
the borderline between a LINER and Seyfert classification.
[OII] H? H? H? [OIII] H? + [NII] H? + [NII]
[OII] H? H? H? [OIII] [OI] [SII]
Keck2 2016 Jun 04
? ?
60 SDSS 2004 Jun 16 150 DCT 2018 Apr 11
SDSS 2003 Apr 07
40 100
20
50
0
0
3000 4000 5000 6000 7000 8000 3000 4000 5000 6000 7000 8000
Rest Wavelength (A?) Rest Wavelength (A?)
(a) iPTF16bco (b) ZTF18aahiqfi
H? + [NII] H? + [NII]
[OII] H? H? H? [OIII] [OI] [SII] [OII] H? H? H? [OIII] [OI] [SII]
60 200
DCT 2018 May 06 ? ? DCT 2018 Jun 21 ? ?
SDSS 2002 Dec 29 150 SDSS 2001 Feb 15
40
100
20
50
0 0
3000 4000 5000 6000 7000 8000 3000 4000 5000 6000 7000 8000
Rest Wavelength (A?) Rest Wavelength (A?)
(c) ZTF18aaidlyq (d) ZTF18aasuray
H? + [NII] H? + [NII]
[OII] H? H? H? [OIII] [OI] [SII] [OII] H? H? H? [OIII] [OI] [SII]
150 P60 2018 Dec 03 DCT 2019 May 02 ? ?
SDSS 2003 Jan 05 SDSS 2007 Feb 18
200
100
100
50
0 0
3000 4000 5000 6000 7000 8000 3000 4000 5000 6000 7000 8000
Rest Wavelength (A?) Rest Wavelength (A?)
(e) ZTF18aasszwr (f) ZTF18aaabltn
Figure 2.4: Comparison of early and follow-up spectra of the other CLAGN in the
sample. Note that the Palomar 60-inch ?P60? spectra have a difference in aperture
affecting the flux measurement by a factor of order unity. Detailed follow-up of
ZTF18aajupnt (not shown here) is presented in Figure 2.6.
Due to its similarity to a TDE at early times, we promptly initiated a multi-
wavelength follow-up campaign of ZTF18aajupnt which we describe in the following
sections. Following the discovery of a blue continuum with the Double Spectrograph
(DBSP) of the Palomar 200-inch Hale telescope on 2018 June 12 (PI: David Cook),
we monitored ZTF18aajupnt with five additional epochs of optical spectroscopy
36
F (10?17? erg/cm2/s/A?) F? (10?17 erg/cm2/s/A?) F (10?17? erg/cm2/s/A?)
F (10?17? erg/cm2/s/A?) F? (10?17 erg/cm2/s/A?) F? (10?17 erg/cm2/s/A?)
Table 2.3: Spectroscopic Legacy and Follow-up Observations of ZTF18aajupnt.
Obs UT Instrument Exposure (s) Reference
2002 July 11 SDSS 28816 Abolfathi et al. (2018)
2018 June 12 Palomar 200? DBSP 2400 This work
2018 July 22 Palomar 60? SEDM 2430 This work
2018 July 30 Swift XRT 40400 This work
2018 Aug 7 Keck LRIS 300 This work
2018 Aug 11 XMM EPIC pn 11906 This work
2018 Aug 12 Palomar 60? SEDM 2430 This work
2018 Aug 12 FTN FLOYDS-N 3600 Arcavi et al. (2018)
2018 Aug 21 Gemini GMOS-N 600 This work
2018 Sept 1 HST STIS 2859 This work
2018 Sept 12 DCT Deveny 2400 This work
with SEDM on Palomar?s 60-inch on 2018 July 22 and Aug 12, LRIS on the Keck
I telescope on 2018 Aug 7 (PI: Kulkarni), Gemini GMOS-N on 2018 Aug 21 (PI:
Hung), and with Deveny on the DCT on 2018 Sept 12 (PI: Gezari). We detail
the configurations of the spectroscopic follow-up observations of ZTF18aajupnt in
Table 2.3. During this time, its optical light curve plateaued in a manner strikingly
similar to iPTF16bco (shown in Figure 2.5). It also surprisingly displayed coronal
emission lines (those detected are shown in Figures 2.7 and 2.21) in a heretofore
low-ionization nuclear source.
Figure 2.6 shows a complete series of spectra obtained for ZTF18aajupnt, as
well as comparisons to some examples of other AGN and transient types, including
the class of extreme coronal line emitters (ECLEs) and the luminous SN IIn SN
2005ip which demonstrated strong coronal line emission (Smith et al., 2009). These
spectra were reduced with standard pipelines and procedures for each instrument.
Measurements of the flux, luminosity, radial velocity, full-width-at-half-maximum
(FWHM), and equivalent width of the emission lines, including the coronal emission
lines ([Fe XIV] ?5304, [Fe VII] ??5721, 6088, [Fe X] ?6376 in the spectrum with the
highest signal-to-noise detection of the coronal lines is given in Table 2.4. The
FLOYDS-N spectrum from 2018 Aug 12 was reported by Arcavi et al. (2018) to
37
22
ZTF18aajupnt/AT2018dyk (r)
ZTF18aajupnt/AT2018dyk (g)
21 ZTF18aasuray (r)
ZTF18aasuray (g)
20 ZTF18aasszwr (r)
ZTF18aasszwr (g)
iPTF 16bco (r)
19 iPTF 16bco (g)
18
17
16
0 50 100 150 200 250 300 350
Days Since Discovery + Constant
Figure 2.5: Difference imaging light curves of the CL LINERs with the best-
sampled P48 observations in the ZTF sample (ZTF18aajupnt, ZTF18aasszwr, and
ZTF18aasuray) plotted in absolute magnitude compared to that of CL LINER
iPTF16bco (triangle shaped points). ZTF18aasszwr and iPTF16bco are similar
in luminosity and more luminous than ZTF18aajupnt and ZTF18aasuray by about
2.5 mag. ZTF18aasuray has a much slower evolution and is constantly redder in
color, whereas ZTF18aajupnt reddens ?280 days into its evolution. The rise of
ZTF18aajupnt mirrors that of iPTF16bco, whereas the decline appears slower than
but similar in shape to that of ZTF18aasszwr.
have broad H?, and both broad and narrow H? and He II. At that time, a blue
continuum was not obvious in their spectrum. However, we show a power-law blue
excess is clearly detected in the residuals of the spectra after subtracting a model
for the host galaxy light (Figure 2.7).
We have corrected for Galactic extinction in the spectra in Figure 2.6, with
color excess E(B ? V ) = 0.0164 mag (from the Schlafly & Finkbeiner (2011) dust
map22). We use the optical correction curve for RV = 3.1 given by Eqs. 3.a. and b.
in Cardelli et al. (1989), such that f = f 10A?/2.5corr obs .
22https://irsa.ipac.caltech.edu/applications/DUST/
38
Absolute Magnitude
[O II] H? H? H? [O III] H? + [N II]
? ?
DCT Deveny 2018 Sept 13
Gemini 2018 Aug 21
FLOYDS 2018 Aug 12
Keck 2018 Aug 7
P200 2018 June 12
SDSS 2002 July 11
ECLE SDSS J0952+2143
ECLE SDSS J0748+4712
SN 2005ip + 138d
NLS1 Mrk 618
iPTF 16bco in Seyfert 1 state
[Fe II]
3000 4000 5000 6000 7000 8000
Rest Wavelength (A?)
Figure 2.6: Host and follow-up spectra of ZTF18aajupnt, alongside various AGN
and coronal line emitters for comparison. AGN emission lines are annotated in gray
and are labeled above the figure. Coronal lines are annotated in red and are labeled
in the middle of the figure. The flux of the H? line (only) in SN 2005ip has been
truncated for visual purposes (as it lies well above the upper boundary of the plot).
Spectra have been rebinned by a factor of four for visual purposes.
2.2.6 UV Imaging and Spectroscopy
We obtained 17 epochs of follow-up imaging of ZTF18aajupnt with the Neil
Gehrels Swift Observatory?s (Gehrels et al., 2004) Ultraviolet/Optical Telescope
39
Normalized F? + Constant
He II?
[Fe XIV]
[Fe VII]
[Fe VII]
[Fe X]
Table 2.4: Line measurements for ZTF18aajupnt from fits in Figure 2.21 and used
in Figures 2.16, 2.13 and 2.17. The blueshift measured significantly only in Fe X
translates to ?0.0005 c.
? F? L vr FWHM EW
(A?) (10?15 ergs s?1 cm?2 A??1) (1039 ergs s?1) (km s?1) (km s?1) (A?)
H? 6562.80 27.67?0.59 82.4?1.2 57? 4 1061? 19 56.9? 1.5
[NII]?6548 6548.05 0.21?0.19 0.64?0.37 ?612? 19 212? 59 0.4? 0.0
[NII]?6583 6583.45 1.11?0.15 3.31?0.29 954? 10 335? 28 7.9? 0.2
H? 4861.30 9.02?0.32 26.85?0.94 76? 8 939? 28 18.0? 0.7
[OIII] 5006.84 0.96?0.16 2.86?0.47 73? 24 489? 59 2.1? 0.3
HeII 4686.00 3.48?0.29 10.37?0.85 10? 28 1157? 69 6.7? 0.6
[FeXIV] 5304.00 0.45?0.14 1.33?0.40 37? 44 546? 115 1.0? 0.3
[FeVII]?5721 5721.00 0.81?0.14 2.40?0.41 62? 40 795? 98 1.6? 0.3
[FeVII]?6088 6088.00 1.08?0.13 3.22?0.39 68? 22 600? 54 2.3? 0.3
[FeX] 6376.00 1.83?0.19 5.44?0.56 ?160? 36 1301? 94 3.9? 0.4
H? + [NII]
[OII] H? H? H? [OIII] [OI] [SII]
1.5 F? ? ??5.46
1.0
0.5
0.0
4000 4500 5000 5500 6000 6500 7000 7500 8000
Rest Wavelength (A?)
Figure 2.7: Host-galaxy-subtracted Keck1 spectrum of ZTF18aajupnt showing pres-
ence of coronal emission lines (red dotted lines) and a blue excess in the residu-
als. We fit the non-stellar blue continuum with a power law (red dashed line with
? = ?5.46).
(UVOT; Poole et al. (2008); Roming et al. (2005)) from 2018 July 30 to 2019 Mar
17 with 2?3 ks per epoch in the UVW2 filter (?eff = 2030 A?; See Figure 2.1 and 2.8).
We detected NUV brightening in the nucleus relative to its archival Galaxy Evolution
Explorer (GALEX; Martin et al. (2005)) All-Sky Imaging Survey (AIS) magnitude
40
Keck1 - SDSS
He II
[Fe XIV]
[Fe VII]
[Fe VII]
[Fe X]
Swift UVW2 Swift XRT
Spitzer 3.6?m ZTF r
Spitzer 4.5?m ZTF g
43.0
42.5
42.0
41.5
41.0
0 100 200 300
Days since 58329.93 MJD
Figure 2.8: The ?L? light curve of ZTF18aajupnt, comparing Spitzer data to con-
current Swift UVOT, XRT and ZTF observations. For the Spitzer and Swift UVOT
observations we subtracted the host galaxy light as estimated by WISE and GALEX
measurements, respectively. To better show the 60-day lag in the X-ray, we fit the
rise caught by optical and X-ray observations with an order 2 polynomial and the
plateau with linear fits.
of NUV = 19.0 mag (measured with a 6 arcsec radius aperture).
The source was initially detected with a Swift UVW2 = 17.7 mag (measured
within a 5 arcsec radius aperture), which then faded to UVW2 = 18.0 mag 20
days later, and then remained roughly at that UV flux over the next 50 days.
Note that while some of the UV flux measured by Swift contains a contribution
from extended star-formation (detected in the UV out to a radius of 15 arcsec),
the fact that it is variable, and brighter than the archival GALEX UV central
flux indicates that it is associated with the transient. The UV-optical color of
ZTF18aajupnt after subtracting off the GALEX flux is UVW2?r = ?0.45 mag,
very similar to iPTF16bco (which had NUV?r = ?0.5 mag, already 0.5 mag bluer
than the color range of AGN in both GALEX and SDSS; Agu?eros et al. (2005);
Bianchi et al. (2005)).
We obtained UV spectroscopy of ZTF18aajupnt with the Space Telescope
Imaging Spectrograph (STIS) FUV and NUV Multi-Anode Microchannel Array
(MAMA) detectors aboard the Hubble Space Telescope (HST) for a 2 ks exposure
41
log ?L?
with 0?.?2 slit width, and G140L (? = 1425 A?) and G230L (? = 2376 A?) gratings
on 2018 Sept 1, 2019 Jan 18 (only in the FUV23), and 2019 March 3, shown in
Figure 2.9 (Proposal ID: 15331, PI: S.B. Cenko).
] IV] II
I]
N V[ +[O+ ] I II] II]+
[C
y? i IV IV IV e I N I Si I
II
L S N C H [ [ Mg
10?13
10?14
2700 2800 2900
10?15
10?16
1250 1500 1750 2000 2250 2500 2750 3000
Rest Wavelength (A?)
Tbb = 4.40? 104 K Tbb = 3.69? 104 K Tbb = 4.61? 104 K NLS1 Mrk 478
F? ? ??2.66 F? ? ??2.32 F? ? ??2.47 NLS1 Mrk 335
AT2018dyk 2018 Sept 01 AT2018dyk 2019 Mar 3 AT2018dyk 2019 Jan 18
Figure 2.9: HST UV spectrum of ZTF18aajupnt compared to two prototypical
NLS1s. Note the presence of high-ionization lines He II, N V, O IV, and C IV, and
the relative weakness of the low-ionization line Mg II ?2798 in ZTF18aajupnt until
later times.
The high spatial resolution of HST (? 0?.?5) enables better isolation of the
nuclear emission from the host galaxy light. The UV continuum, when masking
the emission lines and correcting for Galactic extinction as in Section 2.2.5, is an
equally good fit to both a blackbody (remaining consistent for both observations
within T = (4.5? 0.3)? 104 K) and a power law with spectral index ? = ?2.6? 0.1
where F? = F
?
?,0? or ?? = ???2 = 0.6, with the continuum F?,0 decreasing in flux
by a factor of 10.7 over 140 days, while the strength of the emission lines remain
23The second HST epoch had no NUV coverage due to losing lock on the guide stars, and was
retaken.
42
F? (erg/cm
2/s/A?)
roughly at the same level. This blackbody temperature is not unusual for TDEs
(e.g. Arcavi et al. 2014; Gezari et al. 2012; Holoien et al. 2016a,b; Hung et al. 2017;
van Velzen et al. 2011), and the power-law index is within the range of UV slopes
observed in quasars (?1.5 < ?? < 1.5; Davis et al., 2007), but steeper than the UV
slopes observed in NLS1s (?2 < ?? < 0; Constantin & Shields, 2003). Figure 2.9
shows similarities of the emission features to HST Faint Object Spectrograph (FOS)
spectra of the prototypical NLS1s Mrk 335 and Mrk 478, noting that compared to
the NLS1s, the UV spectrum of ZTF18aajupnt initially has a weaker low-ionization
line Mg II ?2798, which tends to exhibit weak responsivity in CLAGN (e.g. Gezari
et al. 2017; MacLeod et al. 2016). In the latest HST/STIS epoch, ?6 months after
the optical peak, a broad multi-component Mg II line profile appeared, reminiscent
of recently ?awakened? CLAGN Mrk 590 (Mathur et al., 2018). This suggests that
a light travel time delay, and not low responsivity, is responsible for Mg II being
only marginally detected in the intial observation. This also implies that Mg II is
not co-spatial with the Balmer-line emitting region.
Galactic extinction has been corrected in these spectra in the same way as in
Section 2.2.5, but instead using the UV correction curve for RV = 3.1 given by Eqs.
4.a. and b. in Cardelli et al. (1989).
2.2.7 X-ray
We observed ZTF18aajupnt concurrently with 17 exposures of Swift XRT,
detailed in Table 2.5. The XRT data were processed by the XRT Products Page24
(Evans et al., 2009) using HEASOFT v6.2225. We assessed best-fit models utilizing
?2 statistics and XSPEC version 12.9.1a (Arnaud, 1996). Uncertainties are quoted
at 90% confidence intervals. The XRT light curve in the right panel of Figure 2.10
shows that ZTF18aajupnt is a strongly variable X-ray source, caught rising steadily
24http://www.swift.ac.uk/user_objects/
25https://heasarc.gsfc.nasa.gov/docs/software/heasoft/
43
Table 2.5: Swift UVOT/XRT photometry for ZTF18aajupnt. Corresponds to right
panels of Figures 2.10 and 2.11.
Obs UT UVOT/XRT Exposure times Count rate UVW2 Unabsorbed F0.3?10keV L2 keV L2500 A ?OX
(s) (10?2 s?1) (AB mag) (10?13 erg s?1 cm?2 ) (1040 erg s?1) (1042 erg s?1)
2018 Jul 30 931/941 0.4 ? 0.3 17.72 ? 0.04 1.21 0.56 6.80 -1.15
2018 Aug 12 312/2022 0.8 ? 0.3 17.90 ? 0.06 2.64 1.19 5.70 -1.00
2018 Aug 20 491/3001 1.7 ? 0.3 18.05 ? 0.05 4.43 2.38 5.00 -0.86
2018 Aug 22 298/2252 1.0 ? 0.2 18.03 ? 0.06 2.55 1.37 5.10 -0.96
2018 Aug 27 375/3164 1.6 ? 0.3 17.99 ? 0.06 4.18 2.25 5.30 -0.88
2018 Sep 01 286/2874 2.4 ? 0.3 18.05 ? 0.06 6.48 3.48 5.00 -0.80
2018 Sep 18 807/3006 1.7 ? 0.3 18.10 ? 0.05 5.43 2.51 4.80 -0.85
2018 Sep 23 165/3011 2.5 ? 0.3 18.05 ? 0.08 7.93 3.66 5.00 -0.79
2018 Sep 28 324/1877 2.1 ? 0.4 18.20 ? 0.06 6.47 2.98 4.40 -0.80
2018 Oct 03 353/3149 3.4 ? 0.4 18.08 ? 0.06 10.47 4.83 4.90 -0.74
2018 Oct 08 582/2447 3.1 ? 0.4 18.18 ? 0.05 9.60 4.43 4.40 -0.74
2018 Oct 13 1677/1695 3.4 ? 0.5 18.23 ? 0.04 10.53 4.86 4.20 -0.71
2018 Nov 23 1329/2931 2.8 ? 0.3 18.33 ? 0.05 8.68 4.00 3.90 -0.73
2018 Nov 28 1380/2854 2.3 ? 0.3 18.36 ? 0.05 7.27 3.36 3.80 -0.76
2018 Dec 03 1281/2484 3.8 ? 0.4 18.28 ? 0.05 11.82 5.46 4.10 -0.69
2018 Dec 08 629/2452 4.0 ? 0.5 18.33 ? 0.05 12.54 5.79 3.90 -0.67
2019 Mar 17 191/2874 3.9 ? 0.4 18.55 ? 0.09 12.24 5.65 3.20 -0.64
by an order of magnitude in flux over several months. The coadded spectrum (shown
in the left panel of Figure 2.10) is well-modeled by a power law with a spectral
index of ? = 2.82+0.35?0.26 and assuming a Galactic extinction of NH = 1.76? 1020 cm?2
(computed by the NHtot tool; Kalberla et al. (2005); Schlegel et al. (1998)), with no
intrinsic absorption and an observed flux between 0.3?10 keV of (3.0?0.5) ? 10?13
erg cm?2 s?1.
We then observed ZTF18aajupnt with the XMM EPIC pn camera (Stru?der
et al., 2001) on 2018 Aug 11 for a 12 ks exposure (Observation ID: 0822040701,
PI: S. Gezari). We reduced the data using the XMM-Newton Science Analysis
System (SAS) v16.0 (Gabriel et al., 2004). We extracted products with circular
source and background (source-free) regions with radii of 35? and 108?, respectively.
To mitigate background flaring and maximize SNR, we filtered the photometry
for count rates below 1.75 cts s?1. We also adopted CCD event patterns 0 to 4,
corresponding to single- and double-pixel events. We used XMM Newton EPIC-pn
calibration database files updated as of Sept 2018. We fit the XMM EPIC pn data to
a simple power law with spectral index ? = 3.02?0.15 and only Galactic extinction,
characteristic of a steep soft excess, and consistent with the range of photon indices
44
? broken power law10 2 power law (0.3-2 keV)
power law (2-10 keV)
? Background10 3 Swift XRT
10?4
10?5
10?6
10?7
1.0 10.0
10
5
0
1.0 10.0
Energy (keV)
Figure 2.10: Left panel: XRT spectral fit to a broken power law with soft photon
index ? = 2.82+0.35?0.26 described in Section 2.3.6.4. Right panel: Although a slow rise
is evident at the 0.01 counts s?1 level in the hard band (defined as 1.5?10 keV), the
hardness ratio light curve shows that the X-ray flare is primarily soft, i.e. 0.3?1.5
keV.
?1042
power law XMM
4.0 ?1.4
Background Swift XRT
XMM EPIC pn
10?1 3.5
?1.5
3.0
? 2.5
?1.6
10 2
2.0
?1.7
1.5
10?3
1.00
6 1.0 ?1.8
4 0.5
?1.9
2
0.0
0 0 50 100 150 200
1.00 Days since 58329.94 MJD
Energy (keV)
Figure 2.11: Left panel: The XMM EPIC pn data of ZTF18aajupnt fit to a simple
absorbed power law with spectral index ? = 3.02?0.15 shows a prominent, steep
soft excess. Right panel: The X-ray luminosity derived from a power law fit with
?=3 is plotted in comparison with ?OX (described in Section 2.3.6.4).
observed for NLS1s (? = 2.8 ? 0.9; Boller et al., 1996; Forster & Halpern, 1996;
Molthagen et al., 1998; Rakshit et al., 2017).
Using the PIMMS count rate calculator26, the conversion factor between counts
and unabsorbed flux is 3.1?10?11 for XRT, and 1.5?10?12 for XMM.
26https://heasarc.gsfc.nasa.gov/cgi-bin/Tools/w3pimms/w3pimms.pl
45
normalized counts s?1 keV?1 cm?2 normalized counts s?1 keV?1 cm?2
Ratio [Data/Model] Ratio [Data/Model]
L ?10.3?10keV (erg s )
?OX
2.2.8 Infrared
Spitzer Infrared Array Camera (IRAC; Fazio et al. 2004) observations were
triggered for five epochs on 2018 Aug 13 under the approved ToO program (PI:
Yan, PID:13251). At each epoch, the data were taken for both 3.6 and 4.5?m,
each with a total of 600 seconds exposure time. A 50 point cycling dither pattern
was used. The first three epochal data were taken and used for the analysis when
this paper was prepared. The coadded and mosaiced images were produced by the
standard Spitzer pipeline and are directly used by our analysis.
We measured a maximum increase of 0.14 mag compared to archival WISE
observations. We correct the difference magnitude for the small difference between
the bandpass of the two instruments: 0.19, 0.03 mag for channels 1 and 2, respec-
tively, as measured using stars in the field. In Figure 2.8, we show that this ?L?
at 3.6/4.5 ?m (subtracting our estimate of the host galaxy baseline as measured by
WISE) is greater than ?L? in the UV, suggesting a large dust covering factor (the
fraction of solid angle from the central source obscured by dust).
NEOWISE data (WISE, Wright et al. 2010) showed there was no variability
from the host galaxy of ZTF18aajupnt for 1 year prior to its discovery in ZTF,
despite the hint of optical variability observed in June 2016 by iPTF (Section 2.2.2).
2.2.9 Radio
We measure an archival FIRST VLA survey intensity upper limit (including
CLEAN bias) of 0.89 mJy beam?1 at the location of the host of ZTF18aajupnt in
1997.
46
2.3 Analysis
2.3.1 Host Galaxy Classification
We compare the SDSS spectra of the LINER hosts, observed more than a
decade prior to the changing looks caught by ZTF, with follow-up observations taken
using the Palomar 60-inch (P60) telescope and the DCT in Figure 2.4. We fit stellar
absorption and narrow emission lines to the host spectra with pPXF and results are
in Figure 2.12. To distinguish them from star-forming galaxies, Kauffmann et al.
(2003) define a galaxy as a Seyfert if
log([OIII]/H?) > 0.61/(log([NII]/H?)? 0.05) + 1.3.
and Kewley et al. (2001) demarcate a Composite galaxy if
0.61/(log([NII]/H?)? 0.47) + 1.19 < log([OIII]/H?)
is true. These functions are represented as the dashed and solid lines (respectively)
in the BPT [OIII]/H? versus [NII]/H? narrow-line diagnostic diagram shown in
the upper left panel of Figure 2.13. Figure 2.13 also shows various other line ratio
diagnostic diagrams involving the line ratios [OIII]/H?, [NII]/H?, [OI]/H?, and
[OIII]/[OII] (Baldwin et al. 1981; Kewley et al. 2001, Kauffmann et al. (2003),
Kewley et al. 2006), including the WHAN diagram (Cid Fernandes et al., 2011),
accounting for the equivalent width of H? and the fact that the typical BPT LINER
classification contains both ?weak AGN? and ?retired galaxies? that have ceased star
formation.
Analysis of the archival SDSS spectra of the individual sources in this sample
finds that all but CLQ iPTF16bco exist in the borderline region between LINER
47
H  + [NII] H  + [NII]
[OII] H H H  [OIII] [OI] [SII] [OII] H H H  [OIII] [OI] [SII]
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
4000 4500 5000 5500 6000 6500 4000 4500 5000 5500 6000 6500
Rest Wavelength (?) Rest Wavelength (?)
(a) ZTF18aajupnt (b) ZTF18aahiqfi
H  + [NII] H  + [NII]
[OII] H H H  [OIII] [OI] [SII] [OII] H H H  [OIII] [OI] [SII]
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
4000 4500 5000 5500 6000 6500 4000 4500 5000 5500 6000 6500
Rest Wavelength (?) Rest Wavelength (?)
(c) ZTF18aaidlyq (d) ZTF18aasuray
H  + [NII] H  + [NII]
[OII] H H H  [OIII] [OI] [SII] [OII] H H H  [OIII] [OI] [SII]
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
4000 4500 5000 5500 6000 6500 4000 4500 5000 5500 6000 6500
Rest Wavelength (?) Rest Wavelength (?)
(e) ZTF18aasszwr (f) ZTF18aaabltn
Figure 2.12: The SDSS spectra of the host galaxies were fit using the Penalized Pixel-
Fitting (PPXF) method by Cappellari & Emsellem (2004). Red denotes the stellar
population template, blue the emission line fits, and green points the residuals to
the total best fit model. Note the poor fit to the [O II] and [O III] emission lines of
ZTF18aaidlyq, which are replaced in subsequent analysis by the emission line fits
in Figure 2.18. We do not re-analyze iPTF16bco (not shown here) and instead use
the analysis from Gezari et al. (2017).
48
Relative Flux Relative Flux Relative Flux
Relative Flux Relative Flux Relative Flux
1.5 1.5 1.5
Seyfert Seyfert Seyfert
1.0 1.0 1.0
0.5 0.5 0.5
0.0 0.0 0.0
?0.5 ?0.5 ?0.5
SF LINER SF LINER SF LINER
?1.0 ?1.0 ?1.0
?1.0 ?0.5 0.0 0.5 1.0 ?2.0 ?1.5 ?1.0 ?0.5 0.0 ?1.0 ?0.8 ?0.6 ?0.4 ?0.2 0.0 0.2
log([SII]/H?)
2.0 1.5
Seyfert Seyfert ZTF18aajupnt (AT2018dyk)
ZTF18aahiqfi
1.5 1.0
ZTF18aaidlyq
ZTF18aasszwr
1.0 0.5 ZTF18aaabltn
SF ZTF18aasuray
0.5 weak AGN 0.0
0.0 ?0.5
SF LINER
?0.5 ?1.0
retired galaxy (fake AGN)
?1.0 ?1.5
?1.0 ?0.5 0.0 0.5 1.0 ?2.0 ?1.5 ?1.0 ?0.5 0.0
log([NII]/H?) log([OI]/H?)
Figure 2.13: Narrow-line diagnostics for the CL LINER sample in the ?off? state
(i.e. their host galaxies), including iPTF16bco (values from Gezari et al. (2017)).
The majority of the sample is on the borderline between a LINER and Seyfert
classification. Note differences in scale. Upper limits are used when lines are not
significantly detected.
Lower left panel: AGN diagnostic diagram from Cid Fernandes et al. (2011). Only
three of the sources in the CL LINER sample require a Seyfert to power the Balmer
emission lines in their low state, also indicated by the H? line profiles requiring
broad components, shown in the fits in Figure 2.18.
and Seyfert classifications for all five diagnostics shown in Figure 2.13. According
to the diagram of Cid Fernandes et al. (2011), both weak and ?fake? AGN scenarios
are plausible within the 1? errorbars for three LINERs in this sample, excluding
the host of iPTF16bco, which is considered a retired galaxy in this diagnostic, and
the hosts of ZTF18aasszwr and ZTF18aaabltn, which are Seyfert-like (see lower left
panel of Figure 2.13).
We note that the broad H? component of ZTF18aaabltn is not completely gone
49
log(WH?)[A?] log([OIII]/H?)
log([OIII]/[OII])
in the spectrum representing its ?off? state. Although it passed the sample selection
criteria of being identified as a LINER in the Portsmouth SDSS DR12 catalog (de-
scribed in Section 2.2.1), re-fitting of the line ratios of ZTF18aaabltn reveals that it
is a Seyfert rather than a LINER. As we measured a broad base in H?, we classify
it instead as a Sy 1.9 (this is also consistent with prior radio and X-ray detections of
this source). ZTF18aasuray displayed double-peaked broad Balmer emission indica-
tive of an persistent broad line region with unchanging kinematics in both its low
and high states. As the peaks did not represent high enough velocities or asymmet-
ric enough profiles to require separate components, we fit a single broad Gaussian
base in this source when measuring the narrow line ratios. Unlike ZTF18aaabltn,
those measurements were in agreement with the LINER classification.
Similarly to this work, Thomas et al. (2013) also used pPXF to fit stellar kine-
matics and the [S II]/H? ratio diagnostic from Schawinski et al. (2007) (upper right
panel of Figure 2.13) to classify a source as a LINER; however, they used the Gas
and Absorption Line Fitting code (GANDALF v1.5; Sarzi et al. (2017)) to fit emission
lines, whereas we use a simple multi-component Gaussian profile fit to the narrow
lines in the stellar-template-subtracted spectra (see Figure 2.18 in the Appendix for
these model fits). There may also be a discrepancy stemming from GANDALF correct-
ing for dust?the majority of this sample have Balmer decrement fH?/fH? > 3.1,
indicative of strong intrinsic reddening. However, we choose not to apply a dust
correction since it is an uncertain measurement for this sample, due to the weak
emission line intensities.
2.3.2 Black Hole Masses
In order to shed light on the physical differences between the individual AGN
in this sample, we estimate the black hole masses of the CLAGN hosts using several
methods. The broad H? line is the most common virial estimator for BH masses at
50
low redshift (z . 0.4; e.g. Marziani & Sulentic 2012).
( )( )2
? 5 RBLR FWHM(H?)MBH,vir = 1.5 10 M
light days 103kms?1
where R = 32.9( ?L5100A 0.7BLR 44 ?1 ) light days (Kaspi et al., 2000). We also calculate10 ergs
MBH from the host galaxy luminosity following McLure & Dunlop (2002) such that
MBH,Mr = ?0.5Mr,host ? 2.96,
and the host bulge stellar mass using the relation from Ha?ring & Rix (2004)
log(MBH,Bulge[M]) = log(0.0014MBulge[M]),
and computed and from the stellar velocity dispersion (?F) measured from the SDSS
spectrum using the pPXF method) using the MBH ? ? relation from Tremaine et al.
(2002)
logMBH,?F[M] = 8.13 + 4.02log(?
?1
F/200 km s ).
The results of these measurements are summarized in Table 2.2, and discussed
further in Section 2.3.3.
2.3.3 Comparison to Tidal Disruption Events
It is important to compare the properties of this class of AGN ?turning-on?
from quiescence with a related phenomenon of tidal disruption events (TDEs).
When a star passes close enough to a central black hole to be ripped apart by
tidal forces, roughly half of the stellar debris will remain bound to the black hole
and provide a fresh supply of gas to accrete onto the black hole. The evolution of
the flare of radiation from a TDE is regulated by the fallback timescale (tfb), the
51
time delay for the most tightly bound debris to return to pericenter after disruption,
and the circularization timescale, which is dependent on the efficiency at which the
debris streams shock and circularize due to general relativistic precession. Interest-
ingly, the virial black hole mass for all the CL LINERS in the iPTF/ZTF sample
are above the black hole mass for which a solar-type star can be disrupted outside
the event horizon (M . 108BH M). The only exception is ZTF18aajupnt, which
as a NLS1 in its ?on? state, thus with narrower lines, naturally implies a smaller
black hole mass with this method. However, the black hole mass inferred from the
host galaxy velocity dispersion and bulge mass suggest a larger black hole mass of
log(MBH/M) = 7.6?7.8. This trend of the black hole mass from the virial method
being much smaller is consistent with the work of Rakshit et al. (2017), who suggest
that the smaller Balmer line widths measured in NLS1s which lead to lower BH
masses are due to the geometrical effects of being viewed more face-on (?i? = 26?)
compared to normal broad line Sy 1s (?i? = 41?). This claim is backed up by spec-
tropolarimetric studies of NLS1s (Baldi et al., 2016). Alternately, Marconi et al.
(2008) suggested that in rapidly accreting objects (including NLS1s), enhanced ion-
izing radiation pressure could also lead to underestimates of virial black hole mass
estimates.
It is also possible that these transitioning AGN do not obey the radius-
luminosity relation established from reverberation mapping studies of Seyfert galax-
ies. If we instead use the black hole mass estimates from the host galaxy velocity
dispersion, luminosity, and/or stellar mass, we find that the CL LINER sample have
black hole masses of log(MBH/M) ? 7? 8, close to, but not necessarily exceeding
the upper mass limit for the tidal disruption of a solar-type star.
We can also compare the light curves and spectra of our CL LINERs to TDEs.
The quiescence in the light curves before the onset of their flaring activity, their
blue colors (g ? r < 0) during the flaring in most of the cases, as well as their
52
smooth decline from peak are generally consistent with the TDE scenario. The
main distinction is in their spectral properties at peak. The five objects caught
transitioning from a LINER to a type 1 AGN show spectra in their ?on? state that
are almost indistinguishable from normal quasars, besides the relative weakness of
[O III]. In contrast, TDEs exhibit exclusively broad emission lines; broad He II ?4686
emission, and/or broad H? and H? lines, and sometimes broad He I, but with line
luminosities of . 1041 ergs s?1 (Arcavi et al., 2014; Brown et al., 2016; Holoien et al.,
2018; Hung et al., 2017), well below the CL LINERs (see Figure 2.14). Furthermore,
the X-ray spectra of the CL LINERs with X-ray observations in their ?on? state,
iPTF16bco and ZTF18aajupnt, are well described by a power-law, with ? = 2.1
(Gezari et al., 2017) and ? = 3.0, respectively, and are clearly distinct from the
extremely soft blackbody spectra with kT ? 0.04? 0.10 keV characteristic of both
optically and X-ray selected TDEs (Komossa, 2002; Miller et al., 2015; van Velzen
et al., 2019). We present a more detailed comparison of the observed properties of
ZTF18aajupnt with TDEs in Section 2.3.6.
2.3.4 Comparison to Seyfert CLAGN
We measure the H? and [O III] ?5007 luminosities for this sample in their
?on? state in Figure 2.17 and compare to that of SDSS Sy 1s (including NLS1s;
Mullaney et al. (2013)) and quasars (Shen et al., 2011). All AGN in this sample
display [O III] ?5007 luminosities significantly below average for their observed
broad H? luminosity in their ?on? state, consistent with the findings of Gezari et al.
(2017), that CLQs with appearing (disappearing) broadlines were in general closer
to the fringe (average) of the quasar distribution. However, for ZTF18aasszwr and
ZTF18aaabltn, only upper limits of [O III] were possible due to the low SNR for
narrow lines of the low-resolution (R ? 100) follow up spectra.
53
ZTF18aajupnt (AT2018dyk)
iPTF 16bco
ZTF18aaidlyq
ZTF18aahiqfi
ZTF18aasszwr
ZTF18aaabltn
ZTF18aasuray
CLQs (MacLeod 2018)
10
1 10 100 1000
fH?,high/fH?,low
Figure 2.14: Ratio of continuum flux change as a function of broad line flux change
for our changing-look LINER sample (filled shapes) in comparison to changing-look
Seyferts. ZTF18aajupnt (purple circle), is intermediate in flux ratio and H? ra-
tio space between Seyfert CLAGN (black, lower left) and the other CL LINERs in
this sample. The red dotted line denotes a 1:1 ratio between the continuum and
H? fluxes. iPTF16bco, ZTF18aasuray, ZTF18aasszwr, iPTF16bco are outliers in
differential continuum space (although we collected spectra of the latter two with
an IFU spectrograph that can be unreliable at bluer wavelengths), and iPTF16bco,
ZTF18aaidlyq, and ZTF18aahiqfi are outliers in H? luminosity space compared to
that of the Seyfert CLAGN. All have much larger (by a factor of > 10) changes in
broad line flux than the changing-look Seyfert sample. The f?3240 ratio measure-
ments are represented as lower limits, as there is stellar contamination in the low
(LINER) state. Adapted from Figure 6 in MacLeod et al. (2019).
54
f?3240,high/f?3240,low
MacLeod et al. (2019) systematically obtained spectra for highly-variable can-
didate CLQs (defined as type 1 AGN transitioning to type 2s or vice versa) within
the SDSS footprint, requiring Pan-STARRS 1 variability exceeding |?g| > 1 mag
and |?r| > 0.5 mag. We find agreement with their measured positive correlation
between broad emission line and continuum flux changes, but find that our sample
of CL LINERs is more extreme in the parameter space of continuum and H? flux
ratios (ranging from 2?800 and 12?400, respectively) than the CLQ sample from
MacLeod et al. (2019) (with fhigh/flow = 1 ? 7 and 2?8 for continuum and H?,
respectively), shown in Figure 2.14. When restframe flux at 3240 A? was not avail-
able to us due to inconsistent spectral coverage, we measured flux at the shortest
available comparable wavelength.
2.3.5 Eddington Ratio Estimates
We compute the Eddington ratio (Lbol/LEdd) for the sample in their ?on? state
assuming Lbol = 9?L5100A (Kaspi et al., 2000), summarized in the final column of
Table 2.2. Lbol in the ?on? state is measured using difference imaging in the filter
with central wavelength closest to rest-frame 5100 A? for each source (r-band for
higher-redshift sources iPTF16bco, ZTF18aaidlyq, and ZTF18aasszwr, and g-band
for all others). Lbol in the ?off? state is measured from the reddening corrected
L[O III] narrow line luminosity correlation to L2?10 keV for type 2 AGN (Equation 1
in Netzer et al. (2006)) and using the same bolometric correction as Elitzur et al.
(2014), Lbol = 15.8LX .
While virial black hole masses based on the broad H? line and continuum
luminosity are more generally used for AGN, those relations are based on reverber-
ation mapping studies which were never done specifically for NLS1s. Thus, for the
remainder of this work, we adopt BH mass estimates for the sample to be consis-
tent with MBH from stellar velocity dispersions as described in Section 2.3.2 and
55
summarized in Table 2.2.
The Seyfert CLAGN with appearing broad emission lines in the variability-
selected MacLeod et al. (2019) sample (summarized in Section 2.3.4) have ?2 .
log(L/LEdd) . ?1, slightly below that of a control sample of extremely variable
quasars and normal SDSS DR7 quasars. For the relatively large range of this small
sample (?1.7 . log(L/LEdd) . ?0.3), it is difficult to distinguish which popula-
tion?s Eddington ratios to which they are better matched in their ?on? state. The
corresponding upper limits of log (L/LEdd) < ?2 in the ?off? states of the LINER
host galaxies are in good agreement with that of the MacLeod et al. (2019) CL
population that has dimmed.
Elitzur et al. (2014) predict a natural sequence within the disk-wind scenario
in which AGN evolve from displaying to lacking broad optical emission lines. This
evolution is driven by variations in accretion rate (with the critical value param-
2/3
eterized by Lbol/MBH ), as well as the availability of ionizing radiation from the
central engine. The BLR is therefore posited to be assembled following an increase
in accretion rate (likely due to instabilities to match the fast timescales observed;
Rumbaugh et al. (2018)). Due to an insufficient cloud flow rate and lack of ionizing
photons, no BLR can be sustained below the critical accretion rate or bolometric
luminosity (Lbol ? 5 ? 39 2/310 M erg s?17 , Elitzur & Ho (2009)). This spectral evo-
lutionary pathway is supported by modeling an SDSS-selected sample of Seyferts
of various types and spanning L/LEdd ? 10?3 to 0 (Stern & Laor, 2012), for which
accretion rate progressively decreased with luminosity from type 1s to type 2-like
AGN. In Figure 2.15 we recreate this sequence represented by AGN with different
spectral classifications occupying distinct regions of the Lbol?MBH?L/LEdd param-
eter space and roughly separated by the critical threshold of Elitzur & Ho (2009).
We overplot the CL LINER sample in their ?on? states which overlap roughly with
the Seyfert type 1 sources, and in the ?off? LINER states which overlap largely with
56
Quasars (Shen et al. 2011)
46 NLS1s (Mullaney et al. 2013)
Type 1s (Winter et al. 2012)
Type 1.2/1.5 (Winter et al. 2012)
44 Type 2s (Ho 2009)
ZTF18aajupnt (AT2018dyk)
iPTF 16bco
42 ZTF18aaidlyq
ZTF18aahiqfi
ZTF18aasszwr
40 ZTF18aaabltn
ZTF18aasuray
38
4 6 8 10 ?6 ?4 ?2 0
log MBH[M] log Lbol/LEdd
Figure 2.15: AGN, when separated by spectral classification, show the rough evolu-
tionary sequence in parameter space of black hole mass MBH, bolometric luminosity
Lbol, and Eddington ratio Lbol/LEdd described in Section 2.3.5. The dotted lines
denote the critical values above which a BLR can be sustained from Elitzur & Ho
(2009) described in Section 2.3.5, to which we compare the measured values for the
sample (filled, purple shapes with same mapping as in Figure 2.14) and their hosts
(unfilled, black). The CL LINER sample in their ?on? state is consistent with the
type 1s (orange points), with ZTF18aajupnt in its high state (filled circle) toward
the high luminosity end of the NLS1 distribution (blue points). We note that the
type 2 sample from Ho (2009) contains LINER2s and LLAGN. Adapted from Figure
1 in Elitzur et al. (2014).
the type 2s and border on the intermediate type 1.2/1.5 Seyferts.
The bolometric luminosities (and therefore the Eddington ratios) are upper
limits in Figure 2.15 due to the ?off? spectra being almost entirely host dominated.
iPTF16bco, ZTF18aasuray, ZTF18aaidlyq, and ZTF18aasszwr approach the quasars
in their ?on? states, and ZTF18aajupnt does not fall squarely among the NLS1s
but instead in the border region between types. The least luminous sources in
the sample, ZTF18aajupnt and ZTF18aasuray, approach most closely the critical
Eddington ratio for the existence of a BLR in their ?off? states, and the most
luminous iPTF16bco is closest to the intermediate types in its LINER state.
57
log L ?1bol [erg s ]
2.3.6 ZTF18aajupnt: A LINER Changing-Look to a NLS1
For the following analysis we focus on ZTF18aajupnt, for which we have the
most extensive follow-up data, and which showed the appearance of coronal lines
along with X-ray variability. The difference imaging light curve of this event displays
a plateau similar to that of iPTF16bco (Gezari et al. (2017); see comparison in
Figure 2.5), before fading gradually over several months in a manner similar to that
of CL LINER ZTF18aasszwr, rather than the power-law decline characteristic of an
optical TDE light curve (e.g. Hung et al. 2018).
The lack of IR variability in NEOWISE leading up to the turn-on of ZTF18aajupnt
constrains the presence of any IR AGN activity or dust echo in this host to < 10
months. W1-W2 is never greater than ?0.02 during this time, far below the 0.8
threshold AGN diagnostic value from Stern et al. (2012). Stability in the CRTS
light curve similarly confirms that no AGN-like variability was present for 13 years
prior to its discovery with ZTF. There was, however, a hint of some ?0.1 mag flaring
in the CRTS light curve in June 2006 and April 2007. Additionally, we extracted
forced photometry (Masci et al., 2017) for ZTF18aajupnt from the PTF database
covering June 2011 to June 2016, and there were only 8 marginal detections near
the limiting magnitude of PTF (from 20 to 20.9 r-band mag) for the last 15 days of
this range.
To reproduce the photometry of ZTF18aajupnt, any physical explanation must
explain a rise time of ?50 days and a slow decline rate of ?0.5 mag in 60 days, both
quite unusual for a TDE or supernova (e.g. van Velzen et al. 2019). Arcavi et al.
(2018) note that the difference imaging light curve of ZTF18aajupnt peaks at an
absolute magnitude of ?17.4 mag, which is much fainter than the majority of TDEs
by several magnitudes, excluding iPTF16fnl (Blagorodnova et al., 2017). A power
law and blackbody give nearly identical fits to the UV spectra (with T 4bb = 4.5?10
58
K) with no change in the slope as the continuum fades over ?140 days; Figure 2.9).
The optical continuum in Figure 2.7 is well fitted with a power-law, consistent with
the Rayleigh-Jeans tail of a blackbody.
In the UV, the observed spectrum does not resemble that of a TDE in a
LINER (e.g. ASASSN-14li, Cenko et al. (2016)). Instead, the UV spectrum of
ZTF18aajupnt is very similar to the UV spectra of normal NLS1s, with a sim-
ilar spectral slope and peaked, broad emission line shapes (see Figure 2.9). In
particular, ZTF18aajupnt has a strong C IV ??1548, 1551 line and C III] ?1909
line, which is typical of NLS1s, but not detected in all the TDEs with HST UV
spectra: ASASSN-14li (Cenko et al., 2016), iPTF15af (Blagorodnova et al., 2019),
iPTF16fnl (Brown et al., 2018), AT2018zr (Hung et al., 2019). Interestingly though,
ZTF18aajupnt does show N IV] ?1486 emission, which is just barely detected in
NLS1s (Constantin & Shields, 2003) and is detected in the UV spectrum of TDE
ASASSN-14li, which was argued to be N-rich. The critical density 3.4? 1010 cm?3
of the intercombination N IV] ?1486 line provides an upper limit to the density of
this gas in ZTF18aajupnt (Nussbaumer & Storey, 1979). The late-time increase in
the Mg II line has not been detected in a TDE; in fact the opposite trend has poten-
tially been observed: the brightening of broad Mg II with the fading of the transient
in TDE AT2018zr (Hung et al., 2019). Finally, ZTF18aajupnt demonstrates none
of the broad absorption features seen in the UV spectra of TDEs, and has been
associated with powerful outflows launched by the accretion process in a TDE.
2.3.6.1 Coronal Line Emission from ZTF18aajupnt
We report line measurements of the Keck spectrum of ZTF18aajupnt in Ta-
ble 2.4. We choose to analyze the spectrum from this instrument because of its
sufficiently high SNR and spectral resolution to measure the presence of coronal
lines. For each of these measurements, the stellar population of the host galaxy
59
represented by the ppxf fit has been subtracted (See Figure 2.12 for a visual of the
stellar model template).
The width of the majority of the coronal lines is narrower than the widths
of the broad permitted AGN emission lines (see Figure 2.16), as is expected from
forbidden high-ionization collisionally-excited emission because it originates from a
larger distance from the ionization source. However, there is no strong evidence that
the coronal emission lines in ZTF18aajupnt are observed with widths between the
BL and NL emission, as expected in the scenario in which gas is outflowing from an
intermediate coronal line region (CLR; e.g. Mullaney & Ward 2008). The [Fe X]
line is unlikely to be broader than expected due to blending with the [O I] ?6364
line (e.g. Pelat et al. 1987), as it is in a 1:3 ratio with the [O I] ?6300 line which
is observed to be weaker than [Fe X] in this source. In Sy 1s, [Fe X] tends to be
relatively stronger than the other coronal lines (e.g. Pfeiffer et al. 2000). However,
in Seyferts the CL emission is typically measured to be only a few percent of the
strength of [O III] ?5007).
The fact that [Fe X] ?6374 is stronger than [O III] ?5007 places ZTF18aajupnt away
from other Seyferts and instead among the <10 known extreme coronal line emit-
ters (ECLEs) in this parameter space. We discuss further the ECLE scenario in
Section 2.4.2. We note that the weakness of [O III] may be due to light travel time
effects, and thus may strengthen with time.
We note the significant spectral differences between ZTF18aajupnt and SN
2005ip post-peak (Smith et al., 2009). SN 2005ip has much more prominent coronal
lines than even the example ECLEs, as well as a strong hydrogen emission series,
much broader than the quasar iPTF16bco plotted alongside it.
Korista & Ferland (1998) presented a model by which coronal lines are the
result of ISM interaction with bare Seyfert nuclei, i.e. AGN lacking any X-ray/UV
evidence of intrinsic absorption by ionized gas along the line of sight to the AGN.
60
Gemini 2018 Aug 21 FLOYDS 2018 Aug 12 Keck 2018 Aug 7
H? [O III] [N II] H? [N II]
1400
1200
1000
800
600
400
200
0
4750 5000 5250 5500 5750 6000 6250 6500
Rest Wavelength (A?)
Figure 2.16: FWHM of H?, H?, and the coronal lines for each high-resolution optical
observation of ZTF18aajupnt in its ?on? state. The stellar population of the host
galaxy has been subtracted.
This model is consistent with our finding of no intrinsic absorption in the X-ray
spectra of ZTF18aajupnt.
2.3.6.2 ZTF18aajupnt as a NLS1 in its ?On? State
At the other extreme of eigenvectors of AGN spectral properties are narrow-
line Seyfert 1 galaxies (NLS1s), a subclass of AGN that are characterized by rela-
tively narrow Balmer lines (FHWM < 2000 km s?1), strong broad Fe II emission,
[O III] ?5007/H?tot < 3, a prominent soft X-ray excess (e.g. Puchnarewicz et al.
1992), and dramatic variability, especially in the X-rays (e.g. Frederick et al., 2018;
Pogge, 2000). These spectral properties of NLS1s are attributed to lower-mass cen-
tral black holes (5 < log(MBH[M]) < 8; e.g. Mathur et al. 2001) that are thought
61
FWHM (km s?1)
He II
[Fe XIV]
[Fe VII]
[Fe VII]
[Fe X]
F (erg/cm2? /s/A?)
to accrete at high Eddington ratios (Grupe et al., 2010; Pounds et al., 1995; Wang
et al., 1996; Xu et al., 2012).
We measure 1000 . FWHM(H?) < 2000 km s?1 which is indicative of a
narrow-line Seyfert 1 galaxy in the AGN interpretation (Goodrich, 1989), as well as
the fact that the Balmer lines are significantly better fits to Lorentzian line profiles
than Gaussians (Niko lajuk et al., 2009). However, the FWHM limits between Sy 2s,
NLS1s and Sy 1s is somewhat arbitrary (Mullaney & Ward, 2008; Ve?ron-Cetty et al.,
2001), and may even be better set at 2200 km s?1 (Rakshit et al., 2017). The fact
that some of the line measurements fall short of this cutoff could speak to the in-
termediate nature of this transitioning object in the changing-look scenario. The
virial mass measurement for ZTF18aajupnt is consistent with the NLS1 interpreta-
tion, as NLS1s display properties consistent with AGNs with lower masses (Grupe
& Mathur, 2004), though it is toward the high end of the NLS1 mass distribution
(Xu et al., 2012). Also consistent with the NLS1 scenario is that [O III] ?5007 /
H? = 0.1 < 3 (Osterbrock & Pogge, 1985). However, [O III] ?5007 appears to
be relatively quite weak when compared to that of of prototypical NLS1, Mrk 618,
in Figure 2.6. It should also be noted that the coronal lines in ZTF18aajupnt ap-
pear to be symmetric and at the same systematic redshift as the Balmer series
and low-ionization forbidden lines, whereas coronal lines in Seyferts can be signifi-
cantly broadened, asymmetric, and blueshifted consistent with an outflowing wind
launched between the BLR and NLR (Rodr??guez-Ardila et al. (2006); in NLS1s:
Erkens et al. (1997); Mullaney & Ward (2008); Porquet et al. (1999)). This is less
common, but not unheard of, for ECLEs (See Section 2.4.2).
It is evident from all follow-up spectra of ZTF18aajupnt in Figure 2.6 that
it is also missing the prominent Fe II pseudo-continuum complex characteristic of
NLS1s. Therefore we do not utilize an Fe II template in subsequent optical nor UV
spectral fitting. The intense ionizing radiation and high temperatures inferred from
62
the presence of the coronal line emission should make visible the multiply ionized
Fe II were it present. The fact that Fe II lags behind H? in reverberation mapping
studies of AGN (Barth et al., 2013) could mean that not enough time has passed
for this component to be irradiated, consistent with the weak presence of [O III]
(Figure 2.17). Runnoe et al. (2016) also found that, for some CLAGN, the Fe II
complex was only present in the ?on? state. In AGN there is a robust negative
correlation between [O III] and Fe II (the so-called Eigenvector 1; Boroson & Green
(1992)), which manifests typically as weak [O III] in NLS1s (e.g. Rakshit et al. 2017),
possibly indicating we should expect Fe II to become stronger in ZTF18aajupnt after
the light-travel delay time.
44
iPTF 16bco
43 ZTF18aasszwr
ZTF18aahiqfi
42
ZTF18aaidlyq
ZTF18aaabltn
ZTF18aajupnt
41
ZTF18aasuray
40
39 40 41 42 43
log ([O III] Luminosity (erg/s))
Figure 2.17: H? and [O III] ?5007 line luminosities measured for this sample of
CL LINERs, in the high state. The upper and lower contours representing log LH?
vs. log L[O III] measurements of SDSS DR7 quasars and Sy 1 galaxies from Shen
et al. (2011) and Mullaney et al. (2013) show that this sample is up to an order of
magnitude underluminous in [O III], due to light-travel time delays of an extended
narrow line region that has yet to respond to the continuum flux change. The lower
limits of ZTF18aasszwr and ZTF18aaabltn are due to the [O III] ?5007 emission
line not being resolved in the low-resolution (R ? 100) follow-up spectra. Adapted
from Figure 6 in Gezari et al. (2017).
63
log (H? Luminosity (erg/s))
Narrow He II is frequently observed in AGN, however we measure strong He II
broader than the Balmer emission lines (Figure 2.16), possibly revealing an inner
nuclear region not typically probed by the Balmer emission lines alone. This has
been seen in a number of Seyferts such as the Sy1 Mrk 509, but is far less common.
He II ?1640 and [C III] ?1909 observed in the UV spectrum are consistent with
the presence of higher ionization coronal lines in the optical. All prominent emission
features are similar in strength and width to those in the HST FOS spectrum of
NLS1 Mrk 335 and Mrk 478, shown in Figure 2.9 for comparison, however, with
a Mg II ?2798, which is only marginally detected in the first HST/STIS epoch,
and then brightens significantly 4 months later. However, like [O III], the late-time
brightening of Mg II is likely a result of light travel time delays if the Mg II and
[O III] line emitting gas resides further out from the central black hole.
2.3.6.3 The Accretion Rate of ZTF18aajupnt
The Eddington ratio of ZTF18aajupnt ranged between 0.02 and 0.09 from 2018
June to September, assuming the average BH mass of log MBH[M]=7.1 (estimates
described in Section 2.3.2 from stellar velocity dispersion as well as from the standard
virial method; Shen et al. 2011). Note that we assume a constant for the bolometric
correction, but the SED is likely changing throughout the evolution of this source
given the dramatic variability in ?OX described below. This L/LEdd is toward the
low end of the NLS1 distribution, and on the high end for that of CLQs (MacLeod
et al., 2019; Xu et al., 2012). The range of Eddington ratios for the remainder of
the sample is 0.03?0.8. ZTF18aajupnt is probing a critical inflection point in ?OX
and Eddington ratio space related to accretion rate driven state changes analogous
to that of X-ray binaries (Ruan et al., 2019a).
64
2.3.6.4 X-ray Light Curve and Spectra of ZTF18aajupnt
We initially measure a soft X-ray luminosity of a few ?1041 erg s?1 from the
first Swift XRT observations of ZTF18aajupnt on 2018 July 30. Wang et al. (2011)
require at least a few?1042 erg s?1 to power the CLR, a level which ZTF18aajupnt did
not reach until ?40 days later. The XRT light curve in the right panel of Figure 2.10
shows that ZTF18aajupnt is a variable X-ray source (we note that high-amplitude
X-ray variability is characteristic of NLS1s; e.g. Niko lajuk et al. 2009). The excess
variance (or fractional amplitude of variability) defined by Nandra et al. (1997) as
?
?2 = 1 Nrms i=1(xi? x?)2??x2 of the 0.3-10 keV 130-day light curve27 is 0.41, similarNx?
to that of the most variable NLS1s, but high for Sy 1s (Grupe et al., 2000). We
measure a maximum luminosity of LX = (3.7?0.4)?1042 erg s?1. This X-ray lumi-
nosity is difficult to obtain with even the brightest supernova explosions, which have
been observed up to ? 1041 erg s?1 (Immler & Lewin, 2003), and it is toward the
lower end for both Seyferts and NLS1s (Hasinger, 2008). The hardness ratio light
curve in the right panel of Figure 2.10 shows that the X-ray flare is primarily in the
soft band i.e. 0.3?1.5 keV, while the 1.5?10 keV light curve tracks the variability
but with a much smaller amplitude. In contrast, the optical and UV photometry
displays a plateau during this time, reminiscent of that of iPTF16bco (Figure 2.1,
2.5), before declining over several months in a manner similar to ZTF18aasszwr.
The simultaneous optical-to-X-ray spectral slope ratio (?OX) defined as
?OX = 0.3838 log(L2 keV/L2500A)
by Eq. 4 of Tananbaum et al. (1979), and Eq. 11 of Grupe et al. (2010), over
several epochs, measures roughly how an object?s SED is changing with time, and
is strongly correlated with Eddington ratio (Poole et al., 2008). However, Grupe
27The detections used to compute the excess variance were in units of counts.
65
et al. (2010) argue that this correlation is only a reliable estimator for Eddington
ratio for sources with ? ? 1.6, above which the relationship saturates. We derive
?OX from the multi-epoch concurrent observations between 2018 July 30 (61 days
after discovery) and Dec 08 by Swift XRT and UVOT (taken with the UVW2 filter,
which has a central wavelength of 1928A? and FWHM 657A?; Poole et al. (2008)).
ZTF18aajupnt shows dramatic variability in the X-rays (rising by an order of
magnitude in 5 months with LX that varied between (0.4? 3.1) ?1042 erg s?1; see
Figure 2.10). However, the range of ?OX for ZTF18aajupnt (-1.15?-0.67) is fairly
consistent with that of typical LINER values (?1.4 < ?OX < ?0.8; Maoz (2007))
and systematically shallower than that of Type 1 Seyferts (?2.0 < ?OX < ?1.2
(Elvis et al., 1994; Steffen et al., 2006)) and most NLS1s (?1.8 < ?OX < ?0.9;
Gallo (2006).
The soft X-ray spectrum and coronal line emission in ZTF18aajupnt are shared
characteristics with NLS1s. The soft X-ray component in excess above the extrapo-
lation of hard X-ray power-law continuum is observed in a large fraction of Seyfert
AGN (Singh et al., 1985), but is particularly strong in NLS1s. The full extent of
the soft excess component remains unknown, and its origin is debated. It has been
ruled out as the tail of the UV thermal emission from the accretion disk (Gierlin?ski
& Done, 2004; Miniutti et al., 2009; Piconcelli et al., 2005; Porquet et al., 2004) but
Comptonization of those seed photons by an optically thick medium is now one of
the favored scenarios (e.g. Done et al. 2012), as is blurred ionized disk reflection
Garc??a et al. (2019).
Due to their high ionization potentials (? > 100 eV), coronal lines can probe
the soft X-ray excess indirectly, as well as the SED in the vicinity of 200 eV, which
is difficult to observe otherwise because of both Galactic and intrinsic photoelectric
absorption, but important due to their significant contribution to Lbol. Erkens et al.
(1997) found that coronal lines were more likely to be present in Seyferts with
66
steeper X-ray spectra. Gelbord et al. (2009) found in a sample of Seyfert galaxies a
correlation between soft X-rays and [Fe VII], [Fe X], and [Fe XI] lines, proposed by
Murayama & Taniguchi (1998a,b) to originate from the innermost wall of the dusty
torus (see also Rodr??guez-Ardila et al. (2002)).
NLS1s also display strong coronal line emission (e.g. Stephens 1989). Optical
coronal lines include the forbidden transitions of iron, [Fe XIV] ?5304, [Fe VII]
?6088, [Fe X] ?6376 and [Fe XI] ?7894, as well as [Ar XIV] ?4414 and [S XII]
?7612. The coronal lines in NLS1s can be blueshifted with asymmetric velocity
profiles and broad wings, consistent with an outflow (Erkens et al., 1997; Nagao
et al., 2000; Porquet et al., 1999). Gelbord et al. (2009) found [Fe X]/[O III] to
be the most extreme (by a factor of 2-3) in NLS1s with the narrowest broad lines
(FWHM(H?)?800 km s?1) during a search for AGN with strong coronal lines in
SDSS, and interpreted these sources as having strong soft excesses.
2.4 Discussion
While the number of CLAGN is steadily increasing, there has yet to be a large-
scale systematic study of newly-discovered candidates that simultaneously tracks the
appearance of continuum variability and the broad-line emission in real-time using
high-cadence difference imaging photometry.
The best-studied target-of-interest in this sample was identified from ZTF
based on its TDE-like rise time, and therefore we obtained several epochs of support-
ing data in real-time throughout its evolution. Its months-long plateau, UV/optical
spectra, and high-energy properties were indicative of having changed look to a
NLS1. Although they are typically highly X-ray variable, such dramatic optical
variability of a NLS1 has only been seen in three other sources to-date, including
CLAGN NGC 4051 (Guainazzi et al., 1998; Uttley et al., 1999), although it changed
from an obscured Sy 2 and not a LINER. ZTF18aajupnt is therefore unique not only
67
among this sample, but among CLAGN overall.
2.4.1 A New Class of changing-look LINERs
We establish this particular class of CLAGN associated with extreme order-of-
magnitude changes in continuum and emission line flux compared to less dramatic
changing looks occurring in Seyferts (shown in Figure 2.14).
Although most CLAGN reported to-date are Seyferts, this may be due to
sample selection bias, as the high numbers of LINERs may cause them to be seen as
galaxy contaminants in such searches. Difference imaging offers a unique mechanism
to discover variability in known LINERs.
2.4.2 Is ZTF18aajupnt a TDE or AGN activity?
We focus specifically on ZTF18aajupnt, which shows the appearance of broad
Balmer and coronal lines within 16 years of being spectroscopically confirmed as a
LINER, accompanied by an order-of-magnitude soft X-ray flare. Given a ROSAT
All-Sky Survey flux upper limit of F < 5 ? 10?13 ergs s?1 cm?20.1?2.4 keV at the
location of the host from 1990 to 1991 (Voges et al., 1999), ZTF18aajupnt has
therefore displayed a changing look in both the optical and X-ray usages of this
term. The lower limit for this change in soft X-ray flux (0.1-2.4 keV) was by a
factor of 7 at the time of the most recent observation.
Although highly photometrically variable on their own, flares due to non-AGN
mechanisms are not unheard of in NLS1s. For example, CSS100217:102913+404220
displayed a high state (MV = ?22.7 at 45 days post-peak) accompanied by broad H?
and was interpreted either as a Type IIn SN (Drake et al., 2011) or TDE (Saxton
et al., 2018) near the nucleus (?150 pc) of a NLS1. It eventually faded back to
its original level after one year. PS16dtm (or iPTF16ezh/SN 2016ezh) was a ?
1.7?104 K, and near-Eddington but X-ray-quiet nuclear transient with strong Fe II
68
emission which plateaued over ?100 days while maintaining a constant blackbody
temperature. The event was interpreted as a TDE exciting the BLR in a NLS1
(Blanchard et al., 2017), although Oknyansky et al. (2018) claimed it may instead
be a CLAGN transitioning into a Sy 1. No X-rays were observed during follow-
up, dimming at least by an order of magnitude compared to archival observations,
but were predicted to reappear after the obscuring debris had dissipated. SDSS
J1233+0842 was discovered as a CLQ when it changed into a composite type galaxy
or transition object (with [O III]/H? = ?0.10 and [N II]/H? = ?0.17 from Figure
2.a. in MacLeod et al. (2019)). It shows variable Fe II emission (similar to PS16dtm),
with the broad line emission disappearing between 2005 and 2016.
A nuclear transient in the nearby ULIRG F01004-2237 was classified as a
TDE?despite an unusually long peak time of 1 year?partially based on the strength
of its He II compared to H? (Tadhunter et al., 2017). This ratio was unprecedented
for AGN activity, even for AGN in the high state of a changing look. We note that
although it is broad, He II/H? ? 0.4 for ZTF18aajupnt is far below that measured
for F01004-2237. It was later argued that the nature of this transient may instead
be due to changes in the accretion flow, similar to that of OGLE17aaj, which also
showed a slow optical rise and long plateau and slow decline and UV and X-ray
properties similar to that of ZTF18aajupnt, although it lacked spectral classifica-
tion prior to discovery of the transient (Gromadzki et al., 2019). The transient
AT2017bgt was classified as a dramatic SMBH UV/optical flare which irradiated
the BLR and was interpreted as the result of increased accretion onto the SMBH
(Trakhtenbrot et al., 2019a). Unlike ZTF18aajupnt, it showed no decrease in flux
over several months. The persistence of the UV emission distinguished it from SNe
and TDEs, and the extremely intense nature of the UV continuum as well as pres-
ence of Bowen fluorescence He II, [N III] ?4640, and [O III] double-peaked features
in the unobscured optical spectrum distinguished it from CLAGN. As in the ?on?
69
state of ZTF18aajupnt, the Balmer FWHM in all 3 sources are consistent with that
of NLS1 galaxies.
ECLEs are most typically thought to be the echoes of TDEs via the accretion
of tidal disruption streams by previously non-active SMBHs (Wang et al., 2012).
However, less than 10 ECLEs have been reported in the literature, most notably
SDSS J0952+2143 (Komossa et al. (2008, 2009); Palaversa et al. (2016); also tech-
nically a NLS1 using the unconventional cutoff in Rakshit et al. (2017), see Sec-
tion 2.3.6.2 for details), and SDSS J0748+4712 (Wang et al., 2011). We confirm
that ZTF18aajupnt is technically an ?extreme? CLE by the definition put forth by
Wang et al. (2012), because the strength of [Fe X] ?6376 is comparable to that of
[O III] ?5007, as well as by the presence of [Fe XIV] in the optical spectrum (seen in
Figures 2.6 and 2.21) following the independent definition of Palaversa et al. (2016).
We note, however, that it is the present weakness of [O III] that is driving this diag-
nostic, and the coronal lines overall do not appear nearly as strong when compared
to the prototypical ECLEs, SDSS J0952+2143 and J0748+4712, in Figure 2.6. This
strong, slowly variable transient nuclear coronal line emission necessitates soft X-ray
flaring outbursts from an accretion disk, which may be formed as tidal debris set-
tles, illuminating the outermost debris as well as intervening ISM (Komossa & Bade,
1999). The coronal lines in these sources, some blueshifted, faded on timescales of
1-5 years, with strong [O III] appearing even later. Because strong coronal line
emission is not a TDE diagnostic in isolation, some ECLE galaxies with persistent
coronal lines may instead be Seyferts.
IC 3599 is an optical changing-look (displaying dramatic variability in not
only Balmer lines but also [Fe VII] and [Fe XIV]) Sy 1.9 galaxy with strong soft
X-ray repeating outbursts from its galactic nucleus which can be modeled by a
disk instability with a rise time of ?1 year whereby the inner disk is vacated and
subsequently refills (Brandt et al., 1995; Campana et al., 2015; Grupe et al., 1995,
70
2015; Komossa & Bade, 1999). It is the only AGN which has shown fading of its
coronal lines (though this variability is common among non-active ECLEs).
The Swift/XRT and XMM spectra of ZTF18aajupnt fit well to a steep power
law (? ?3?0.2) below 2 keV, not a disk blackbody as would be expected in the
TDE scenario (see Figures 2.10 and 2.11). The large covering factor measured
for ZTF18aajupnt by Spitzer is also more consistent with mid-infrared studies of
CLAGN (Sheng et al., 2017), than the covering factor derived for TDEs with dust
echoes (with fdust = Edust/Eabsorb at the ?1% level; van Velzen et al. 2016). This
could imply appreciable accretion happening recently, because that is very likely
required for a dusty torus with a large covering factor. In an accretion event un-
related to disk physics, a self-gravitating molecular cloud with low enough angular
momentum could also be efficiently accreted on the correct timescales, activating
radiation which subsequently illuminates the BLR (e.g. Hopkins et al. 2006). One
way to obtain a larger covering factor would also be via chaotic cold accretion, by
which interaction via inelastic collisions is made easier, boosting the funneling of
molecular clumpy clouds toward the SMBH, and therefore enhancing the accretion
rate. (Gaspari & Sadowski, 2017). The high blackbody temperature measured from
UV spectroscopy implies the line of sight to the transient is not significantly dust
obscured. Sheng et al. (2017) argue that mid IR light echoes of CLAGN (with
?W1|W2 & 0.4 mag) was additional evidence to support the reprocessing scenario
driven by changing accretion rate instead of variable obscuration. W1 ? W2 for
that sample varied between 0.1 and 1.2 mag, so [3.6] ? [4.5] ?m = 1.4 mag for
ZTF18aajupnt was consistent with the lowest end of that sample for mid IR color
(it would not have been selected based on its variability amplitude for the short
duration of the Spitzer observations reported here).
LINERs may have inefficient accretion disks surrounding a low-luminosity
AGN, occupying a unique physical parameter space compared to other CLAGN.
71
Similar to the unification scheme derived for AGN (Antonucci, 1993; Urry & Padovani,
1995), broad- and narrow-line LINERs can be categorized into LINER1s and LINER2s
(e.g. Gonza?lez-Mart??n et al. 2015; Ho et al. 1997a,b). Yan et al. (2019) reported the
discovery of the ?turning on? of a type 1 Seyfert occurring in LINER SDSS1115+0544 which
flared for ?1 year and subsequently plateaued, followed by a mid-IR dust echo de-
layed with respect to the optical by 180 days and a late-time UV flare, although
no soft X-rays were detected then. Narrow coronal lines appeared in the spectrum
along with H? and H? consistent with broad line emission. As was done in Yan
et al. (2019), we measured the soft X-ray-[Fe VII] ratio for ZTF18aajupnt to be log
L2 keV/L[Fe VII]?6088 =1.25 at maximum, still significantly below the average of 3.37
and pointing to an X-ray deficit compared to normal AGN (Gelbord et al., 2009),
although we note that the soft X-rays changed by a factor of 10 and likely contin-
ued to rise beyond our last Swift observation. We also measure a minimum L/LEdd
equivalent to that of SDSS1115+0544. Yan et al. (2019) concluded an instability was
required to ?turn on? an AGN from a quiescent galaxy within hundreds of days.
They argued that (despite a rate in tension with the AGN duty cycle) given the
discovery of iPTF16bco and SDSS1115+0544 one year apart, such events should
not be uncommon, a prediction this sample supports. There must be a connection
between the LINER hosts and the state that is enabling these rapid transitions.
2.4.3 The nature of the high-ionization forbidden ?coronal? lines in
ZTF18aajupnt
Noda & Done (2018) posited that in the well-studied changing-look AGN Mrk
1018, the coming and going of the soft X-ray excess (the main ionization source)
drives the appearance and disappearance of the BLR and therefore the changing-
look phenomenon. We observe strong soft X-rays increasing in luminosity over time,
which are required to form the coronal lines, although we note that the peak of the
72
X-ray flaring appears to lag behind the UV/optical flaring.
The nuclear outburst in UV and X-ray required is similar to cataclysmic vari-
able or black hole binary thermal-viscous disk instability flares, which have been
discussed as a possible mechanism for powering optical changing-looks, although
the observed time scales are much faster than predicted (e.g. Lawrence 2018; Ross
et al. 2018; Siemiginowska et al. 1996; Stern et al. 2018).
Ross et al. (2018) attribute changing looks to a thermal (cooling) front prop-
agating inward through the accretion disk or disk surface opacity changes, which
have the correct timescales for observed transitions, unlike other proposed CLAGN
mechanisms.
We posit that this quiet LINER suddenly goes into an active outbursting state,
the rise in ionizing radiation at first confined to the innermost BLR, turning on into
a NLS1, then flash ionizing the ambient gas in the CLR, whereas the NLR (where
[O III] lines are formed) is at larger distances, and thus light-travel time effects delay
their response. Mg II, though still broad, is formed further out on average (Cackett
et al., 2015; Goad et al., 1993; O?Brien et al., 1995).
2.4.4 The nature of the soft X-ray excess during the NLS1 state of
ZTF18aajupnt
The preceding interpretation does not explain the soft X-ray rise, which is
clearly delayed at least ?60 days with respect to the end of the UV/optical rise
(shown in Figure 2.8), and may speak instead to a lag in an ?outside-in? sense
following the direction of an accretion flow, rather than photon propagation from
a central ?lamp post?. This is in contrast to the clear inter-band time lags on the
order of days in support of the reprocessing scenario measured by Shappee et al.
(2014) in high cadence multiwavelength observations of CLAGN NGC 2617, which
transitioned from a Sy 1.8 to a Sy 1 in 10 years. The ?2 month lag observed in
73
ZTF18aajupnt also suggests that this delay is not simply from light-travel time.
X-ray inter-band time delays in NLS1s measured via Fourier based spectral timing,
due to either X-ray reverberation or propagating fluctuations, are typically on the
order of tens to hundreds of seconds (e.g. Kara et al. 2016; Uttley et al. 2014).
This delayed X-ray response may tell us something fundamental about the
origin of the soft X-ray excess in AGN in general. The long delay of the soft X-
ray flare relative to the expected light-travel time delays between the UV/optical
emitting accretion disk and the compact, hot corona suggests that we are witnessing
the real-time assembly of the corona plasma itself, possibly due to structural changes
due to the dramatic change of state in the inner accretion disk (Garc??a et al., 2019).
If the Balmer emission is indeed from a BLR, we predict the H? and H?
lines should get broader as the UV luminosity decreases. Continued spectroscopic
monitoring to look for evolution in line widths and strengths, particularly the narrow
[O III] emission line and Mg II, and monitoring of the soft X-rays will be critical to
map out the structure of this system and distinguish between the scenarios presented
here.
2.5 Conclusions
We present the changing looks of six known LINERs caught turning on into
type-1-like AGN found in Year 1 of the ZTF survey. It is the first systematic study of
its kind performed in real time using difference imaging variability as the discovery
mechanism for selecting nuclear transients in these previously quiescent galaxies.
1. We establish a class of changing-look LINERs, distinct from Seyfert changing-
look AGN, with unique spectroscopic and photometric variability properties
intrinsically due to the LINER accretion state.
2. In their ?on state? the changing-look LINERs have suppressed narrow [O III]
74
line emission compared to normal AGN of the same broad H? luminosity, and
inferred Eddington ratios 1?4 orders of magnitude above their LINER state.
3. This sample includes a multiwavelength study between 2018 June to 2019
March of the first case of a LINER changing look to a NLS1 ? ZTF18aajupnt ?
which transitioned within 3 months based on its archival light curve.
4. We observed the delayed response of the NLR and broad Mg II with respect to
the appearance of broad (yet < 2000 km s?1) Balmer lines, and X-ray flaring
delayed ?60 days with respect to the optical/UV rise of this nuclear transient,
indicative of an ?outside-in? transition.
5. We interpret this particular object to be a dramatic change of state in a pre-
existing LINER accretion disk, which eventually forms an optically thick inner
structure that up-scatters the UV/optical seed photons to produce a delayed
soft X-ray excess.
This class of previously-weak AGN has the potential to be a laboratory with
which to map out the structure of the accretion flow and surrounding environment.
We plan to continue to monitor the behavior of these transients, and expect to build
upon the sample at a rate of ?4 year?1 for the next two years of the ZTF survey.
2.6 Appendix
75
?10?16 ZTF18aajupnt (AT2018dyk) ?10?17 ?10?16
3 5.0 1.5
2.5
2 1.0
0.0
0.5
1
?2.5
0.0
0
?5.0 ?0.5
6520 6540 6560 6580 6600 4800 4825 4850 4875 4900 4925 4925 4950 4975 5000 5025 5050 5075
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
? ?16 ZTF18aahiqfi ?10?17 ?10?1610
1.5 4 1.0
2
1.0 0.5
0
0.5 ?2 0.0
0.0 ?4
?0.5
6520 6540 6560 6580 6600 4800 4825 4850 4875 4900 4925 4925 4950 4975 5000 5025 5050 5075
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
ZTF18aaidlyq ?10?17 ZTF18aaidlyq
10.0
10 7.5
2
5.0
5
2.5
0
0 0.0
? ?2.52
?5 ?5.0
6500 6525 6550 6575 6600 6625 4800 4825 4850 4875 4900 4925 4950 4975 5000 5025 5050 5075
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
? ?16 ZTF18aasszwr ?10?17 ?10?1710
4
1.5
7.5
1.0 2 5.0
2.5
0.5 0
0.0
0.0 ?2
?2.5
6450 6500 6550 6600 6650 6700 4800 4825 4850 4875 4900 4925 4950 4975 5000 5025 5050 5075
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
? ?16 ZTF18aaabltn ?10?16 ?10?1610
6
3 1.0
4
2 0.5
2
1 0.0
0
0 ?0.5
6450 6500 6550 6600 6650 6700 4800 4825 4850 4875 4900 4925 4940 4960 4980 5000 5020 5040 5060 5080
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aasuray ?10?16 ZTF18aasuray ?10?16 ZTF18aasuray
2
4
0.5
1
2
0.0
0
0
?0.5
6450 6500 6550 6600 6650 6700 4800 4825 4850 4875 4900 4925 4950 4975 5000 5025 5050 5075
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
Figure 2.18: Stellar-continuum-subtracted H? line complex, H?, and [O III] (first,
second, and third columns, respectively) with best fits to Gaussians for the sample
in their ?off? state used in Figure 2.13. While fitting H? in ZTF18aaidlyq only, the
FWHM in the model fit has been fixed to the FWHM of [N II] ?6585, due to H?
being only marginally detected in that host. Note the differences in scale.
76
F? (erg/cm
2/s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?) F 2 2? (erg/cm /s/A?) F? (erg/cm /s/A?)
[N II]?6548
[N II]?6548
[N II]?6549
H?
[N II]?6548 [N II]?6548 H?[N II]?6548
H? H?
H?
H?
[N II]?6583 [N II]?6583 [N II]?6583
[N II]?6585
[N II]?6583 [N II]?6583
F? (erg/cm
2/s/A?)
F? (erg/cm
2/s/A?) F 2? (erg/cm /s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?)
H? H?
H? H?
H? H?
F (erg/cm2? /s/A?)
F? (erg/cm
2/s/A?) 2 2
2 F? (erg/cm
2/s/A?) F? (erg/cm /s/A?) F? (erg/cm /s/A?)
F? (erg/cm /s/A?)
[OIII]
[OIII] [OIII] OIII OIII OIII
?10?16 ZTF18aajupnt (AT2018dyk) ?10?17 ZTF18aajupnt (AT2018dyk) ?10?16 ZTF18aajupnt (AT2018dyk)
4
2 1.0
2
1 0.5
0
0
?2 0.0
?1
3675 3700 3725 3750 3775 3800 6200 6250 6300 6350 6400 6650 6675 6700 6725 6750 6775
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aahiqfi ?10?17 ZTF18aahiqfi ?10?17 ZTF18aahiqfi
6
4
4
1
2
2
0 0 0
?2
?2
?1
3675 3700 3725 3750 3775 3800 6200 6250 6300 6350 6400 6650 6675 6700 6725 6750 6775
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
?10?17 ZTF18aaidlyq ?10?17 ZTF18aaidlyq ?10?17 ZTF18aaidlyq
4
7.5 2
2
5.0 1
0
2.5 0
0.0 ?2?1
?2.5 ?4
?2
3680 3700 3720 3740 3760 6200 6250 6300 6350 6400 6650 6675 6700 6725 6750 6775
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aasszwr ?10?17 ZTF18aasszwr ?10?17 ZTF18aasszwr
1.00
2 4
0.75
0.50 1 2
0.25
0
0
0.00
? ?10.25 ?2
3675 3700 3725 3750 3775 3800 6200 6250 6300 6350 6400 6650 6675 6700 6725 6750 6775
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aaabltn ?10?17 ZTF18aaabltn ?10?16 ZTF18aaabltn
1.00
4 4 0.75
0.50
2 2
0.25
0
0 0.00
?2 ?0.25
3675 3700 3725 3750 3775 3800 6200 6250 6300 6350 6400 6650 6675 6700 6725 6750 6775
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aasuray ?10?17 ZTF18aasuray ?10?16 ZTF18aasuray
1.5
2 2
1.0
1 0 0.5
0
?2 0.0
?1
3675 3700 3725 3750 3775 3800 6200 6250 6300 6350 6400 6650 6675 6700 6725 6750 6775
Rest Wavelength (A?) Rest Wavelength (A?) Rest Wavelength (A?)
Figure 2.19: Stellar-continuum-subtracted [O II], [O I], and the [S II] doublet (first,
second, and third columns, respectively) line profiles and best fits to Gaussians for
the sample in their ?off? state used in Figure 2.13. Note the differences in scale.
77
F? (erg/cm
2/s/A?) F (erg/cm2/s/A?) F (erg/cm2/s/A?) F (erg/cm2? ? ? /s/A?)
F 2 2? (erg/cm /s/A?) F? (erg/cm /s/A?)
[OII]
[OII] [OII] [OII]
[OII] [OII]
F? (erg/cm
2/s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?)
[OI] [OI] [OI] [OI] [OI] [OI]
F (erg/cm2/s/A?)
F? (erg/cm
2/s/A?) ? F? (erg/cm2/s/A?) F (erg/cm2? /s/A?) F? (erg/cm2/s/A?) F? (erg/cm2/s/A?)
[S II]?6716 [S II]?6716 [S II]?6716 [S II]?6716 [S II]?6716[S II]?6716
[S II]?6730 [S II]?6730 [S II]?6730 [S II]?6730 [S II]?6730[S II]?6730
?10?15 ZTF18aajupnt (AT2018dyk) ?10?16 ZTF18aajupnt (AT2018dyk)
6
1.00
0.75 4
0.50
2
0.25
0
0.00
6500 6525 6550 6575 6600 6625 4800 4825 4850 4875 4900 4925
Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aahiqfi ?10?16 ZTF18aahiqfi
2.0
8
1.5
6
1.0
4
0.5
2
0.0
0 ?
6300 6400 6500 6600 6700 6800 4600 4700 4800 4900 5000 5100
Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aaidlyq ?10?16 ZTF18aaidlyq
1.00
3
0.75
2
0.50
0.25
1
0.00
0 ?0.25
6300 6400 6500 6600 6700 6800 4600 4700 4800 4900 5000 5100
Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aasszwr ?10?16 ZTF18aasszwr
3
8
6 2
4 1
2
0
0
6300 6400 6500 6600 6700 6800 4600 4700 4800 4900 5000 5100
Rest Wavelength (A?) Rest Wavelength (A?)
?10?15 ZTF18aaabltn ?10?16 ZTF18aaabltn
2.0
8
1.5
6
1.0 4
0.5 2
0
0.0 ?
6300 6400 6500 6600 6700 6800 4600 4700 4800 4900 5000 5100
Rest Wavelength (A?) Rest Wavelength (A?)
?10?16 ZTF18aasuray ?10?16 ZTF18aasuray
7.5 3
2
5.0
1
2.5
0
0.0 ? ?1
6450 6500 6550 6600 6650 6700 4600 4700 4800 4900 5000 5100
Rest Wavelength (A?) Rest Wavelength (A?)
Figure 2.20: Balmer line profiles and best fits to Gaussians/Lorentzians for the
sample in their ?on? state. Note the difference in scales. See Section 2.2.5 for more
details.
78
F? (erg/cm
2/s/A?)
F (erg/cm2/s/A?) 2
F? (erg/cm
2/s/A?) ? F (erg/cm2/s/A?) F (erg/cm2/s/A?) F? (erg/cm /s/A?)? ?
[N II]?6548
[N II]?6548
H? [N II]?6548 [N II]?6548 H?
[N II]?6548 H? H? H?
H?
[N II]?6583 [N II]?6583
[N II]?6583 [N II]?6583
[N II]?6583
F? (erg/cm
2/s/A?) 2 F (erg/cm2/s/A?)F? (erg/cm /s/A?)
? F? (erg/cm
2/s/A?) F? (erg/cm
2/s/A?) F? (erg/cm
2/s/A?)
H?
H? H?
H? H?
H?
?10?16 ?10?17
2.0 7.5
1.5 5.0
1.0
2.5
0.5
0.0
0.0
?2.5
4625 4650 4675 4700 4725 4750 5280 5290 5300 5310 5320
Rest Wavelength (A?) Rest Wavelength (A?)
?10?17 ?10?16
6 1.0
4
0.5
2
0
0.0
?2
?4 ?0.5
5650 5675 5700 5725 5750 5775 6025 6050 6075 6100 6125 6150
Rest Wavelength (A?) Rest Wavelength (A?)
?10?16
1.0
0.5
0.0
6325 6350 6375 6400 6425 6450
Rest Wavelength (A?)
Figure 2.21: Coronal line profiles and best fits to Gaussians for the ZTF18aajupnt in
its ?on? state. The residual of the spectrum to the stellar continuum was fit. See
Section 2.3.6.1 for more details.
79
F 2? (erg/cm /s/A?) F? (erg/cm
2/s/A?)
F? (erg/cm
2/s/A?)
[FeVII]?5721
HeII
[FeX]
F 2? (erg/cm /s/A?)
F? (erg/cm
2/s/A?)
[FeXIV]
[FeVII]?6088
Chapter 3: A Family Tree of Optical Transients from Narrow-Line
Seyfert 1 Galaxies
The Zwicky Transient Facility (ZTF) has discovered five events (0.01 < z <
0.4) belonging to an emerging class of AGN undergoing smooth, large-amplitude,
and rapidly rising flares. This sample consists of several transients initially clas-
sified as supernovae with narrow spectral lines. However, upon closer inspection,
all of the host galaxies display Balmer lines with FWHM(H?) ? 900 ? 1400 km
s?1, characteristic of a narrow-line Seyfert 1 (NLSy1) galaxy. The transient events
are long-lived, over 400 days on average in the observed frame. We report UV and
X-ray follow-up of the flares and observe persistent UV emission, with two of the
five transients detected with luminous X-ray emission, ruling out a supernova inter-
pretation. We compare the properties of this sample to previously reported flaring
NLSy1 galaxies and find that they fall into three spectroscopic categories: 1) Balmer
line profiles and Fe II complexes typical of NLSy1s, 2) strong He II profiles, and 3)
He II profiles including Bowen fluorescence features. The latter are members of the
growing class of AGN flares attributed to enhanced accretion reported by Trakht-
enbrot et al. (2019). We consider physical interpretations in the context of related
transients from the literature. For example, two of the sources show high amplitude
rebrightening in the optical, ruling out a simple tidal disruption event scenario for
those transients. We conclude that three of the sample belong to the Trakhtenbrot
et al. (2019) class, and two are TDEs in NLSy1s. We also hypothesize as to why
NLSy1s are preferentially the sites of such rapid enhanced flaring activity.
80
3.1 Introduction
A galaxy center hosting an active galactic nucleus (AGN) is dominated by its
continuum emission. Therefore, a flare originating from this nuclear region requires
a distinctly powerful event to be detectable above this stochastically variable con-
tinuum. A small number of rapid1, smoothly evolving flares have been observed to
be associated with AGN (e.g. Blanchard et al. 2017; Drake et al. 2011), with few
known mechanisms that can cause these events to occur.
Intrinsic UV/optical flares, such as those due to enhanced accretion onto the
central supermassive black hole (SMBH) in the form of gaseous material or stars
passing too close to the nucleus, have been observed in the form of: tidal disruption
events (e.g. Gezari et al. 2012; van Velzen et al. 2020a), UV-bright flaring events that
are associated with accretion rate changes (Trakhtenbrot et al., 2019a), transients
with double peaked line profiles linked to accretion disk emission (e.g. Halpern &
Eracleous 1994), or changing-look AGN ? the dramatic change in spectroscopic
AGN classification following a rise in continuum level, thought to be connected to
unstable changes in accretion state (e.g. Frederick et al. 2019; Gaskell & Sparke
1986; Graham et al. 2020; LaMassa et al. 2015; MacLeod et al. 2016; Ross et al.
2018; Ruan et al. 2016; Runnoe et al. 2016; Stern et al. 2018; Trakhtenbrot et al.
2019b).
Phenomena extrinsic to the SMBH accretion engine, such as microlensing of
a quasar by a foreground Galactic source (e.g. Lawrence et al. 2012) or slowly
evolving super-luminous supernova (SLSN) explosions, have also been observed to
cause smooth large-amplitude flares from galaxies with AGN (Graham et al., 2017).
In rare cases these can be astrometrically indistinguishable from the galactic nucleus,
and therefore it becomes difficult to discern whether an explosive disruption to the
1We refer to flare timescales as ?rapid? when they occur on week to month timescales.
81
accretion flow has occurred, and to differentiate this from AGN variability (Terlevich
et al., 1992).
Multiwavelength approaches are required to disentangle this diverse family of
observed flaring behaviors from AGN. In the golden era of time domain astronomy,
even with many multichromatic instruments trained on the sky, a number of newly-
discovered objects continue to defy placement into a clear-cut observational category.
In order of discovery, we present a photometric class comprised of five rapid
flares with similar smooth light curve shapes occurring in a subclass of AGN observed
by the Zwicky Transient Facility (ZTF) survey:
a) ZTF19aailpwl/AT2019brs (z = 0.37362)
b) ZTF19abvgxrq/AT2019pev (z = 0.097)
c) ZTF19aatubsj/AT2019fdr (z = 0.2666)
d) ZTF19aaiqmgl/AT2019avd (z = 0.0296)
e) ZTF18abjjkeo/AT2020hle (z = 0.103)
In Section 3.2 we present the follow up of these flares. In Section 3.3 we
compare the results of their respective multiwavelength follow up campaigns to ob-
servations of a variety of related objects found in recent years, and in Section 3.4 we
attempt to place them into a classification scheme based on observational properties,
summarized in Section 3.5. All transients in the sample are referred to by their ZTF
alert names throughout. All magnitudes are reported in the AB system and light
curves are shown in the observed frame unless otherwise stated. We have adopted
the following cosmology: H = 70 km s?10 Mpc
?1, ??= 0.73 and ?M = 0.27.
82
3.2 Observations
The Zwicky Transient Facility Survey (Bellm et al., 2019a; Graham et al.,
2019) is comprised of the automated Palomar 48-inch Samuel Oschin Telescope
(P48) as well as the Palomar 60-inch SED Machine (P60 SEDM; Blagorodnova et al.
2018; Rigault et al. 2019), and has surveyed the Northern Sky with g- and r-band
filters with a 3-night cadence since 2018 (Bellm et al., 2019b). At least 15 images
meeting good quality criteria were stacked to build a coadded reference image of
each observing field and quadrant in each filter band. Science images are subtracted
by their references and processed each night by the Infrared Processing and Analysis
Center (IPAC) pipeline (Masci et al., 2019). The candidate transient alert stream
(Patterson et al., 2019) is distributed by the University of Washington Kafka system,
and filtered through the AGN and black holes Science Working Group?s Nuclear
Transients2 parameter criteria (outlined in van Velzen et al. 2019, 2020b) by the
Ampel broker (Nordin et al., 2019; Soumagnac & Ofek, 2018), with the GROWTH
Marshal user interface utilized for the coordination of follow-up efforts (Kasliwal
et al., 2019).
All 5 transients included in the sample presented here were selected based on
the following criteria: large amplitude, nuclear variability (?g > 1 mag in difference
imaging photometry, and within 0.5?? of the center of the host galaxy in the reference
image) with follow-up or pre-flare spectra consistent with an AGN classification.
This selection was not systematic (and therefore not complete), but rather the result
of ongoing intersecting and collaborative searches for changing look AGN (Frederick
et al., 2019), TDEs (van Velzen et al., 2019, 2020b), and superluminous supernovae
(Lunnan et al., 2019; Yan et al., 2020) relying on partial human vetting from the
2A nuclear transient was defined as that within 0.5? of the reference galaxy center. Over 9000
nuclear transients passed this filter and were ranked during ZTF Phase I, of which 27 were TDEs,
over 7% were classified as SN, and over half were AGN or candidate AGN.
83
ZTF transient alert stream, from which this sample emerged as more examples were
collected. A systematic search for NLSy1 transients in ZTF will be the focus of a
future study.
3.2.1 Optical Photometry
All transients in the sample were detected pre-peak using ZTF difference imag-
ing photometry. The smooth light curve shapes (with scatter ?g < 0.1 mag) of the
sample are shown in Figure 3.1. All magnitude changes are reported in g band
unless otherwise noted. An analysis of the rise times to peak are measured and
reported in Section 3.3.1.1. We report the g-band magnitude-weighted offsets for
each transient, calculated using Equation 3 in van Velzen et al. (2019). ZTF forced
photometry for the sample is shown in Figure 3.10 of the Appendix.
AT2019brs ? (RA=14:27:46.41, Dec=+29:30:38.6, J2000.0) was first detected
on 2019 Feb 08 as a nuclear transient within 0?.?17 of the host galaxy center. The host
galaxy displayed some variability at the < 1 mag level in the Catalina Real-Time
Transient Survey (CRTS; Drake et al. 2009) from 2005 to 2013.
AT2019pev3 ? (RA=04:29:22.72, Dec=+00:37:07.6, J2000.0), also known as
Gaia19eby, was first detected on 2019 Sept 04 as a nuclear transient within 0?.?15 of
the host galaxy center. ATLAS, Gaia, and PanSTARRs also reported observations
of this source on the Transient Name Server (TNS) with discovery dates of 2019
Sep 04, 2019 Sep 13, and 2019 Sep 26, respectively. The host galaxy displayed no
variability above the 0.5 mag level in CRTS.
AT2019fdr ? (RA= 17:09:06.86, Dec=+26:51:20.7, J2000.0) was detected on
2019 Apr 27 with a significant flux increase with respect to the reference image and
3AT2019pev passed the ZTF TDE working group?s tidal disruption event criteria, and was
given the nickname ?Stannis Baratheon? for ease of discussion. When it was found to be among
a class of AGN-associated objects serendipitously detected by ZTF, the other sources in the class
were retroactively given the names of other Game of Thrones characters in the same Great House
- and collectively referred to fondly as ?The Baratheons?, whose motto is, appropriately, ?Ours is
the Fury?.
84
?26 s s s
AT2019brs/
ZTF19aailpwl - 2.5 mag
?24
AT2019fdr/
?22 sssss s s s s ZTF19aatubsj - 2 mag
sssssss s s s
?20 AT2019pev/
ZTF19abvgxrq + 0.5 mag
?18
s s
? AT2019avd/16 ZTF19aaiqmgl + 1.5 mag
?14 AT2020hle/s ZTF18abjjkeo + 5 mag
0 200 400 600 800 1000
Days Since Discovery
Figure 3.1: Comparison of the ZTF g- and r-band difference imaging light curve
shapes and absolute magnitudes of the sample. AT2019fdr decreases before
reaching a second plateau stage, and undergoes significant reddening after the
first plateau while the others never do. AT2019pev rises again symmetrically after
decreasing to pre-flare levels, as does AT2019avd. The light curves have been
shifted in absolute magnitude space for visual purposes, as indicated alongside the
object name. Overlap of the g and r light curves reflects true colors such that the
initial colors approach g ? r = 0 mag for all transients in the sample. Observations
at other wavelengths are shown in Figure 3.2. Spectroscopic epochs are labeled for
each light curve with an ?S? below AT2019fdr and AT2020hle and above the rest.
with an offset from the nucleus of its host of 0?.?13. During a coverage gap in the
first 40 days of the rise, ATLAS reported an intriguing ?bump? feature (Smartt
et al., 2019). The host galaxy displayed variability at the 2 mag level in V-band
CRTS data from 2009 to 2013 (variability which was not observed in ZTF forced
photometry prior to the transient).
85
Absolute Magnitude
AT2019avd ? (RA=08:23:36.77, Dec=+04:23:02.5, J2000.0), also known as
eRASSt J082337+0423034, was detected by ZTF beginning on 2019 Feb 09 within
0?.?06 of its host galaxy. The host showed no variability in CRTS for 15 years prior
to its rapid rise to peak.
AT2020hle ? (RA=11:07:42.91, Dec=+74:38:02.0, J2000.0) was detected be-
ginning on 2020 Apr 05 within 0?.?02 of its host galaxy center. The ZTF forced
photometry for this source shows no variability above the level of the galaxy for
> 400 days. The host galaxy of AT2020hle was beyond the survey limits of CRTS.
3.2.2 Optical Spectroscopy
All spectroscopic follow-up observations for the sample are summarized in
Table 3.1, and each epoch is shown in Figure 3.11 of the Appendix. The phases of
the optical follow-up spectra with respect to the features in the ZTF light curves are
annotated on Figure 3.1. All transients in this sample have spectral characteristics of
NLSy1 galaxies, i.e. strong Balmer line emission with FWHM < 2000 km s?1, along
with other spectral features which are highlighted below and explored in detail in
Section 3.3.2. We reduced Palomar 60? SED Machine (P60/SEDM; Program PIs:
Gezari, Sollerman, Kulkarni) spectra with pysedm (Rigault et al., 2019), and all
other spectra with pyraf using standard procedures.
AT2019brs ? showed a striking difference to the SDSS spectrum showing
it was a NLSy1 as early as 2006 (Abolfathi et al., 2018; Rakshit et al., 2017).
The follow-up Folded Low Order whYte-pupil Double-dispersed Spectrograph North
(FLOYDS-N; Arcavi et al. 2019 and Lowell Discovery Telescope (LDT, formerly
DCT; PI: Gezari) spectra showed a steep blue continuum and a strong He II profile
with Bowen fluorescence features, indicating it became a flaring SMBH belonging
4This was the only source in the sample to be detected by the extended ROentgen Survey with an
Imaging Telescope Array (eROSITA, part of the Russian-German ?Spectrum-Roentgen-Gamma?
(SRG) mission; Cappelluti et al. 2011), and was given the name eRASSt J082337+042303. This
X-ray detection coincident with the transient?s host galaxy is described in Section 3.2.4.
86
Table 3.1: Summary of spectroscopic follow-up observations of the sample.
Name Obs UT Instrument Exp (s) Reference
AT2019brs 2006 Jul 01 SDSS 3000 Abolfathi et al. (2018)
2019 Mar 15 FLOYDS-N 3600 Arcavi et al. (2019)
2019 Jun 22 LDT Deveny 900 This work
AT2019fdr 2019 May 25 Palomar 60? SEDM 2250 This work
2019 Jun 17 LT SPRAT 900 This work
2019 Jun 22 LDT Deveny 900 This work
2019 Jul 03 Palomar 200? Hale 600 This work
2020 Apr 30 NOT ALFOSC 1750 This work
AT2019pev 2019 Sep 08 Palomar 60? SEDM 2250 This work
2019 Sep 15 LT SPRAT 500 This work
2019 Sep 22 Palomar 60? SEDM 2250 This work
2019 Sep 24 LDT Deveny 600 This work
2019 Sep 25 Keck LRIS 300 This work
2019 Sep 25 NICER 2000 Kara et al. (2019)
2019 Oct 01 Chandra LETG 45400 Miller et al. 2019
2019 Oct 05 Lick 3-m KAST 1500 This work
2019 Oct 12 LT SPRAT 500 This work
2019 Oct 15 Chandra LETG 91000 Mathur et al. 2019
2019 Oct 23 LDT Deveny 900 This work
2019 Nov 01 Palomar 60? SEDM 2250 This work
2019 Dec 03 LDT Deveny 2400 This work
2020 Feb 26 LDT Deveny 2600 This work
2020 Jan 30 Swift XRT 94700 This work
AT2019avd 2020 Mar 15 NOT ALFOSC 1800 Malyali et al. (2021)
2020 Apr 28 eROSITA SRG 140 Malyali et al. (2021)
2020 May 10 FLOYDS-S 3600 Trakhtenbrot et al. (2020)
AT2020hle 2020 May 16 Palomar 60? SEDM 2250 This work
2020 May 18 LT SPRAT 1000 This work
to the observational class established by Trakhtenbrot et al. (2019a).
AT2019pev ? was spectroscopically identified as a NLSy1 on 2019 Sept 15 with
the Liverpool Telescope (LT; PI: Perley) SPectrograph for the Rapid Acquisition
of Transients (SPRAT), based on the width of the Balmer emission lines and the
strength of the [O III] ?5007 emission line. Gezari et al. (2019) reported that
the LT spectrum showed evidence for blue-shifted He II ?4686 emission as well as
87
N III ?4640 emission, due to the Bowen fluorescence mechanism, placing it again
in the observational subclass of the Trakhtenbrot et al. (2019a) objects. Near peak
it was observed with Keck 10-m Low Resolution Imaging Spectrometer (LRIS; PI:
Graham) as well as the LDT Deveny Spectrograph (PI: Gezari) and the KAST
Double Spectrograph on the Lick 3-m Shane Telescope (PI: Foley), which confirmed
the strong blue continuum and clearly defined and persistent Bowen fluorescence
features.
AT2019fdr ? was observed 8 days after peak on 2019 Jul 03 with the Double
Spectrograph (DBSP) on the Palomar 200-inch Hale Telescope (P200; PI: Yan). We
measured a significant ?blue horn? component of H? and marginally detected He II.
The transient continuum of AT2019fdr faded to reveal an underlying Fe II complex
in the Nordic Optical Telescope (NOT; PI: Sollerman) spectrum taken nearly 368
days after peak on 2020 Apr 30, with no evidence for He II emission.
AT2019avd ? The spectrum taken with NOT (PI: Sollerman) on 2019 Mar 15
near the first optical peak showed strong Balmer line emission, no detection of a He II
line complex, and evidence for a Fe II complex, characteristic of NLSy1 galaxies.
A follow-up FLOYDS-S spectrum taken 444 days after peak and reported to the
Transient Name Server (TNS) by Trakhtenbrot et al. (2020) showed the appearance
of He II and Bowen fluorescence features and a ?blue horn? in H?. Again this event
was classified as a member of the Trakhtenbrot et al. (2019a) observational class of
flaring NLSy1s.
AT2020hle ? In the LT (PI: Perley) spectrum of AT2020hle taken on 2020
May 18 8 days after peak, the narrow component of the He II profile is significantly
blueshifted. No Fe II line complex was detected in the spectra of this transient.
88
3.2.3 UV Photometry
We triggered target-of-opportunity monitoring observations with the Neil Gehrels
Swift Telescope (Gehrels et al., 2004) for all transients in the sample. Using the
HEASOFT command uvotsource we extracted UVOT photometry within a 5??-
radius circular aperture and using an annular background region centered on the
coordinates of the optical transient.
Figure 3.2 shows the ?L? light curves of all flares in the sample. We com-
pare ZTF g and r band difference imaging, WISE difference imaging, Swift XRT
monitoring, and Swift UVOT detections subtracted by the archival Galaxy Evolu-
tion Explorer (GALEX; Bianchi et al. 2017) All-Sky Imaging Survey (AIS) near-UV
(NUV, ?eff = 2310 A?) host measurements (measured with a 6-
??radius aperture).
We found all transients in the sample to be UV-bright, but with varying UV
colors. The UV color of AT2019avd (UVW1?g = ?0.2 mag) was similar to that of
AT2019pev (UVW2? g = ?0.2 mag) and AT2019brs (ranging from UVW2? g =
?0.1 mag to ?0.7 mag in 80 days), but AT2019fdr was the only transient in the
sample with positive UV color (UVW2? g = 0.8 mag). The UV light curves of the
sample tend to follow the shape of the optical. AT2019pev became host dominated
over time as the transient faded. but with strong scatter in the light curve as it
approached the host magnitude.
3.2.4 X-rays
We found only two transients in the sample to be X-ray bright in follow-up
Swift XRT observations: AT2019pev and AT2019avd. AT2019brs was detected
only once, and then only at a low level. We measured an XRT upper limit of
0.004 counts s?1 for AT2019fdr. X-ray follow-up spectra are reported in Table 3.1.
Swift photometry compared to WISE W1- and W2-band and ZTF g and r-band
89
photometry is shown in Figure 3.2. The X-ray bright flares in this sample tend to
vary in lockstep with the slow UV/optical flares.
AT2019brs ? was detected only once in 11 observations during a 16-month
monitoring campaign between 2019 Mar 21 and 2020 Jul 7. We measured a 3-?
detection of 0.003 counts s?1 on 18 Apr 2019, just brighter than the limiting flux.
AT2019pev ? Similar to the UV light curve, the shape of the X-ray flare of
AT2019pev followed the optical, from its fade through its second rise (See Figure 3.2
and Section 3.3). The unabsorbed 0.3?10 keV flux from the stacked XRT spectrum
of AT2019pev was 7.3?0.1?10?12 erg cm?2 s?1. AT2019pev was previously detected
in ROSAT, and NICER observations show an increase in flux from this by 100 times
(1? 10?11 erg cm?2 s?1), variable from 11 to 14 counts s?1 in 3 hours (Kara et al.,
2019). A 50 ks Chandra LETG grating observation taken just 8 days after peak and
reported by Miller et al. (2019) found a flux consistent with this, with the spectral
shape a good fit to a kT = 0.24 keV blackbody, and the source variable at the 25%
level on 2 ? 3 ks timescales. Mathur et al. (2019) reported a decrease in 0.3 ? 2.5
keV flux to 7.7 ? 10?12 erg cm?2 s?1; their 91 ks Chandra LETG observation was
a good fit to a consistent blackbody model and a power law component typical of
AGN with spectral index ? = 1.8, with no intrinsic absorption required.
AT2019avd ? was observed only during the second optical flare (on 2020 Apr
28, 350 days after the first ZTF detection), and was the only X-ray bright transient
in the sample with much fainter X-ray ?L? than that of the optical (shown in
Figure 3.2). Like AT2019pev, the shape of the X-ray rise followed that of the
second rise. It was detected by eROSITA as eRASSt J082337+042303, a soft X-ray
transient consistent with the galaxy 2MASX J08233674+0423027 (Malyali et al.,
2020, 2021). Prior to this, the XMM Slew Survey reported a non-detection at the
location of the host galaxy, with an upper limit of < 1.7 ? 10?14 erg?1 s?1 cm?2
assuming kTbb = 100 eV and NH = 3?1020 cm?2. The SRG flux of 1.5?10?12 erg?1
90
s?1 cm?2 was 90 times brighter than this upper limit. No hard X-ray component was
detected above 2.3 keV. No strong short-term variability on hours-long timescales
was detected, and no strong variability was detected between SRG and the 3 Swift
XRT monitoring observations taken afterward with a week-long cadence. Swift and
NICER observations over the next 5 months showed an additional increase in X-ray
flux by a factor of 10 (Pasham et al., 2020). A careful study of the X-ray properties
of this transient is forthcoming (Malyali et al., 2021).
3.2.5 IR
Malyali et al. (2020) reported that the WISE color of AT2019avd was atypically
low (W1?W2 ' 0.07 mag) compared to typical AGN values (W1?W2 = 0.7?0.8
mag, Assef et al. 2013; Stern et al. 2012). The WISE colors of AT2020hle (neoWISE:
0.35 mag, AllWISE: 0.036 mag) and AT2019pev (W1 ?W2 = 0.45 mag) are also
inconsistent with an AGN, though not quite as low as that of AT2019avd. Only
AT2019brs truly appeared as an AGN in IR, with W1?W2 = 0.98 mag. The WISE
AGN classification of the sample is summarized in Table 3.3 in Section 3.4.
A flare in the IR was detected in NeoWISE at the location of AT2019avd and
concurrent with the optical and X-ray transient. Though the IR flare began much
sooner in 2009, Figure 3.2 shows that the peak of the flare was delayed with respect
to the first optical peak. Prior to this flare, WISE photometry detected no variability
at the location of AT2019avd for nearly 5 years.
3.3 Analysis
3.3.1 Photometry
The difference imaging light curves for the sample are shown in terms of ab-
solute magnitudes in Figure 3.3.
91
Swift UVW1 eROSITA SRG ZTF g
Swift UVM2 W1 ZTF i
Swift UVW2 W2 P60 r
Swift XRT ZTF r P60 g
43.0
44.75
44.50 42.5
44.25
44.00 42.0
43.75
41.5
43.50
43.25
41.0
43.00
AT2019brs/ZTF19aailpwl AT2019avd/ZTF19aaiqmgl
42.75 40.5
0 100 200 300 400 500 600 700 800 0 100 200 300 400 500 600 700 800
Days Since Rise to Peak Days Since Rise to Peak
43.4
44.00
43.75 43.2
43.50
43.0
43.25
43.00 42.8
42.75
42.6
42.50
42.4
42.25
42.00 AT2019pev/ZTF19abvgxrq 42.2 AT2020hle/ZTF18abjjkeo
0 100 200 300 400 500 0 50 100 150 200 250 300
Days Since Rise to Peak Days Since Rise to Peak
44.50
44.25
44.00
43.75
43.50
43.25
43.00
42.75 AT2019fdr/ZTF19aatubsj
0 100 200 300 400 500 600 700
Days Since Rise to Peak
Figure 3.2: We track the colors of the transients in the sample with a ?L? light
curve, comparing the ZTF and WISE data to concurrent high cadence Swift UVOT
and XRT monitoring observations. The X-ray rise and fade of AT2019pev tracks
the optical/UV with no significant delay. We subtracted the host galaxy light as es-
timated by GALEX NUV measurements from the Swift UVOT observations. Times
are given in days since first ZTF detection. The X-ray errorbars are comparable to
the size of the data points. See Figure 3.10 for pre-outburst forced photometry.
92
log ?L? [erg s
?1] log ?L? [erg s?1] log ?L ?1? [erg s ]
log ?L? [erg s
?1] log ?L? [erg s?1]
We show the sample alongside various NLSy1-related events from the liter-
ature, which are described in more detail in Section 3.4. CSS100217 displayed
some variability prior to the transient, unlike any of the events in this sample.
AT2017bgt was observed only during its fade in difference imaging, so we instead
show its aperture photometry (from the ASAS-SN Photometry Database5; Jayas-
inghe et al. 2019) which also shows the rise of the source. AT2018dyk is by far the
least luminous transient shown. We note the similarity of the shapes of the light
curves of AT2019fdr and PS16dtm, which is discussed further in Section 3.4.
3.3.1.1 Light Curve Timescales
We measured the rise-to-peak timescales of the sample by fitting Gaussians to
the light curves shown in Figure 3.3 using the lmfit package built from scipy.optimize.
We observe a correlation between the luminosity (specifically the absolute magni-
tudes MV and Mg) and rise-to-peak timescales of the sample (trise) with the following
relation: M = ?0.04trise ? 18.59, shown in Figure 3.4. Fitting the light curves with
quadratic functions resulted in the same correlation within the error estimates. In-
terestingly, TDEs also show a positive correlation between rise time to peak and
luminosity (van Velzen et al., 2020b). AT2018dyk appears under-luminous for how
fast it rises. AT2017bgt was observed only during its fading phase in difference
imaging, and so was excluded from this portion of the analysis.
3.3.1.2 Rebrightening
It is noteable that two sources in the sample, AT2019pev and AT2019avd, each
have a dramatic rebrightening episode. Following a flare and an approximately ?2
mag fade from peak, both return to nearly half their maximum luminosity before
seasonal gaps in visibility. This is in contrast to that of almost all TDEs and
5https://asas-sn.osu.edu/photometry
93
SN in the literature (e.g. Sollerman et al. 2019, 2020), although they can show
plateaus and ?humps? (e.g. Hammerstein et al. 2020, in prep.)6 We explore possible
interpretations of this rebrightening in Section 3.4.
3.3.1.3 UV/Optical to X-ray Ratio
We derive the simultaneous UV/optical-to-X-ray spectral slope ratio (?OX)
from the Swift UVOT and XRT observations of the sample, (as well as upper limits
assuming ?X = 2, when applicable). We compute unabsorbed X-ray flux densi-
ties at 2 keV using the PIMMS v4.10 web tool7. Following Eq. 4 of Tananbaum
et al. (1979) and Eq. 11 of Grupe et al. (2010), the definition of this ratio is
?OX = 0.3838 log(L2 keV/L2500A). Of the transients detected in X-rays, the ?OX of
AT2019pev evolves over 150 days between 1.1 and 1.4, and AT2019brs is observed
in X-rays during only one epoch with ?OX = 1.7, equivalent to that of the late time
detections of AT2019avd. The range of ?OX measured for the sample is consistent
with that of NLSy1s (0.9 < ?OX < 1.8; Gallo 2006).
3.3.2 Spectroscopy
From the FWHM of the broad Balmer emission lines, we classified all sources
in the sample as NLSy1s. We fit the H? and H? line profiles of the host (when
available) and transient spectra of the sample with the non-linear least-squares mini-
mization and curve-fitting routine in the lmfit Python package. The results of these
fits are shown in Figure 3.5. Using a Lorentzian profile for the broad H? component
fit provided an improvement of the fit over that of a Gaussian profile, as would be
expected based on studies of NLSy1s (e.g. Niko lajuk et al. 2009).
We compare the host (when available) and transient spectra of this sample to
6We note that ASASSN-15lh showed a large amplitude ?double-humped? structure in its UV
light curve.
7https://cxc.harvard.edu/toolkit/pimms.jsp
94
other transients in NLSy1s in Figures 3.6 (showing the full wavelength range of the
observations) and 3.7 (rest wavelength 3700 ? 5150 A?, showing clearly the He II,
Fe II, and H? line profiles). In Figure 3.7 we color-code the sample (as well as
these known NLSy1-related transients in the literature) based on the observational
classification scheme we establish in Section 3.4.3, named after the features discussed
in the following sections: ?He II only?, ?He II+N III?, and ?Fe II only?8.
When compared to the newly discovered flaring events to those in the litera-
ture, it is clear that AT2017bgt (Trakhtenbrot et al., 2019a) has a much stronger
He II+N III Bowen fluorescence profile, CSS100217 (Drake et al., 2009) has stronger
narrow emission lines overall, and AT2018dyk (Frederick et al., 2019) has a weaker
blue continuum. The presence and strength of Fe II is uncorrelated with other spec-
troscopic properties of the transients shown. Of the ZTF sample, the transient spec-
trum of AT2019fdr shows the strongest Fe II complex. However, AT2019fdr shows
no strong He II + Bowen fluorescence features while the others in the ZTF sample
do. AT2019fdr and AT2019avd both show offset blue components of H?.
3.3.2.1 Strong He II profiles in AGN?
In the discovery paper for transient ASASSN-18jd, Neustadt et al. (2020)
emphasized the relatively rare nature of strong He II emission in AGN in general,
noting the exceptions in the Trakhtenbrot et al. (2019a) observational class of flares
as well as the rapid changing-look AGN event AT2018dyk (Frederick et al., 2019). A
strong He II line profile is common (but not ubiquitous) in the spectra of TDEs, and
they are typically accompanied by Bowen fluorescence features (e.g. Blagorodnova
et al. 2019; van Velzen et al. 2020b). AT2019avd, AT2019pev, AT2019brs look the
most similar to AT2017bgt spectroscopically. They are spectroscopically classified
as ?He II+N III?-type flares in Figure 3.7.
8We note that although ?only? is used in the categorization naming based on the presence of
spectral features, all have strong Balmer features.
95
3.3.2.2 The Fe II complex
A strong Fe II line complex (blueward and redward of H?+[O III] in optical
spectra, between 4434 A? and 5450 A?) is a distinguishing feature of NLSy1 galaxies.
Reverberation mapping studies of AGN show that the line complex emitting region
is measured farther than the Balmer line emitting region (e.g. Barth et al. 2013;
Rafter et al. 2013). The Fe II complex seen in PS16dtm was interpreted as evidence
of the system being a NLSy1 prior to the onset of the flare. CSS100217 also displayed
a strong Fe II complex and was interpreted as a SN in a NLSy1 (Drake et al., 2011).
TDE AT2018fyk also showed low ionization lines including an Fe II (37,38) emission
multiplet emerging for 45 days during the tidal disruption event, and forms a class
of Fe-rich TDEs along with ASASSN-15oi and PTF-09ge (Wevers et al., 2019).
Therefore, this feature may indicate the presence of an AGN, but is not always
useful in determining the nature of a particular AGN-related flare. For two of the
transients in this sample, whether or not the Fe II complex can be seen in optical
spectra depends on the phase and the continuum brightness of the transient ? for
AT2019fdr it was not observed for 368 days, and for AT2019avd it became no longer
visible during the second rise 444 days after the initial spectrum was taken.
3.3.3 X-rays
There are only two significantly X-ray detected transients in the sample:
AT2019pev and AT2019avd. We show their X-ray spectra in Figure 3.8 fit to power
law models. The third, AT2019brs, was only detected in one epoch and not at a
level that allowed for the signal-to-noise necessary for a spectrum.
The X-ray spectrum of AT2019pev is measured by Swift XRT with a power
law index of ? = 2.99? 0.02, typical of the strong soft X-ray excess observed below
1 keV in NLSy1s (? = 2.8 ? 0.9; Boller et al., 1996; Forster & Halpern, 1996;
96
Molthagen et al., 1998; Rakshit et al., 2017). The spectrum of AT2019pev could
also be explained by a 150 eV blackbody with a ? = 2 power law component and
no intrinsic absorption (Kara et al., 2019). We note that the soft excess observed
in NLSy1s can mimick the blackbody temperatures expected for TDEs (e.g. Boller
et al. 1996).
The spectral index of AT2019pev (? ? 3) was similar to that of AT2018fyk,
interpreted as a TDE with late-time disk formation (Wevers et al., 2019), as well
as AT2018dyk, interpreted as a changing-look LINER ?turning-on? into a NLSy1
(Frederick et al., 2019). The X-ray spectral index of AT2019avd was quite high even
with regard to these events, with ? ? 4? 6.
3.3.4 Black Hole Masses
We measured the black hole masses of the sample using two different methods,
each with important caveats: The virial mass method, which may systematically un-
derestimate BH masses for NLSy1s, and the host galaxy luminosity, which may be
contaminated by the presence of an AGN. The MBH calculated from the host galaxy
luminosity is MBH,Mr = ?0.5Mr,host ? 2.96 following McLure & Dunlop (2002), and
the standard virial method (e.g. Shen et al. 2011) was employed to obtain the virial
black hole masses from FWHM H? reported in Table 3.2. The transient Eddington
ratio estimates depend on the BH masses (MBH) as LEdd = 1.3 ? 1038 (MBH/M)
erg s?1, For each transient in the sample, we report a range of Eddington ratios in
Table 3.2 bracketed by the Eddington ratio measured assuming the virial mass esti-
mate for the BH mass, and the Eddington ratio measured assuming BH mass derived
from the host galaxy luminosity. The range in BH masses, and therefore Eddington
ratios, shown in Table 3.2 is quite large. We estimate statistical and systematic
uncertainties of 0.3?0.5 dex on these mass and Eddington ratio measurements, due
to the typical scatter associated with single-epoch mass scaling relationships as well
97
as the unknown BLR geometry (e.g. Liu et al. 2018a, 2020; Merloni et al. 2015;
Runnoe et al. 2016).
Miller et al. (2019) obtained an independent measurement of the BH mass of
AT2019pev. They measured M 6BH = 3.7 ? 10 M from the observed Chandra X-
ray luminosity (this observation is described in more detail in Section 3.2.4). This
is closer to, but not consistent with, the virial mass estimate, meaning that the
transient may not have been accreting near the Eddington limit at the time of the
X-ray observation.
Table 3.2: Black hole mass measurements of the sample from optical spectra and
host galaxy properties. NLSy1s are typically thought to be lower mass, highly
accreting systems, but we show here that the uncertainty in the mass estimates
generates significant uncertainty in the estimates of the Eddington ratios (described
in Section 3.3.4). Mr,host is the r-band de Vaucouleurs and exponential disk pro-
file model fit magnitude from the SDSS DR14 photometric catalog. The host of
AT2020hle is not in the SDSS footprint, and so we instead use the Pan-STARRS1
r-band Kron magnitude of this source (Chambers et al., 2016).
Name Mr,host ?L5100A FWHMH log M? BH,Mr log MBH,vir L/LEdd
(mag) (1043 erg s?1) (km s?1) [M] [M]
AT2019pev -21.36 5.00 ? 0.04 878 ? 49 7.7 6.4 0.066-1.5
AT2020hle -20.94 2.24 ? 0.02 1199 ?270 7.5 6.4 0.048-0.62
AT2019brs -22.38 42.6 ? 0.8 1050 ? 77a 8.2 7.2 0.17-1.97
AT2019avd -20.35 0.553 ? 0.008 1433 ? 35 7.2 6.1 0.023-0.29
AT2019fdr -21.51 21.9 ? 0.2 1208 ? 57 7.8 7.1 0.24-1.2
a. The FWHM(H?) for AT2019brs agrees with the measurement in Rak-
shit et al. (2017) within the error estimates.
3.4 Discussion
In this section, we rule out possible physical scenarios for each outburst, begin-
ning with core collapse supernovae IIn. We review why the supernova interpretation
was quickly ruled out in favor of a supermassive black hole accretion scenario, and
discuss how many of the characteristics of the objects are consistent with both
98
NLSy1s and TDEs. We compare the available evidence with other scenarios in-
cluding TDEs, extreme AGN variability, and binary SMBHs in detail. We also
discuss NLSy1 galaxies as the preferential hosts for these and other similar events,
and outline a scheme for classifying future events based on the presence of spectral
features.
3.4.1 ?IIn or not IIn??: Preliminary Observational Classification of
the Flare Sample
Identification of the sample presented here occurred with a slew of conflicting
preliminary classifications at early times, which we describe below.
The narrow emission lines in the spectra of some SLSN (Type IIn) are a result
of the highly luminous interaction of supernova ejecta from a massive progenitor
with dense circumstellar medium. Therefore, under special circumstances, nuclear
SNe can look spectroscopically very similar to rapid9 flares from NLSy1s in the
optical (e.g. Moriya et al. 2018). The shapes of the light curves of the transients in
this sample looked rather like those of such supernovae, in the absence of additional
observations. The smoothness of the flares in particular was unique with respect
to typical stochastic AGN variability, and made these transients noteworthy for
allocation of follow-up resources. Therefore, the narrow Balmer features in the
spectra of these transients, coupled with their light curve shapes, left uncertainty
in their early classifications. They could have been either Type IIn supernovae or
NLSy1 AGN, while those with persistent strong He II ?4686 features in their spectra
looked similar to that of TDEs. To illustrate this, Figure 3.9 shows spectra of the
sample alongside a Type IIn SN as well as a TDE with Bowen fluorescence features.
Additional follow-up observations in the UV/X-rays helped distinguish this sample
of transients from SNe.
9With rise times on the order of days to weeks.
99
3.4.2 A Preponderance of Rapid Optical Transients in Narrow-line
Seyfert 1 Host Galaxies
In the Analysis section (?3.3), we compared our sample to data from nuclear
transients in the literature that happened to be hosted in NLSy1 galaxies. In this
and the next sections, we discuss NLSy1s as an interesting AGN subtype, and
observationally classify and link these events to one another on the basis of their
shared host properties.
The narrower broad-line Balmer profiles and high amplitude variability, (es-
pecially in the X-rays, e.g. Frederick et al. 2018; Pogge 2000) in NLSy1s may be
evidence of smaller black hole masses in these systems (5 < log(MBH[M]) < 8; e.g.
Mathur et al. 2001), and/or higher observed accretion rates (Grupe et al., 2010;
Marconi et al., 2008; Pounds et al., 1995; Wang et al., 1996; Xu et al., 2012). The
virial masses derived from spectral measurements of the population may also be
explained with geometrical effects, when interpreted as the classic broad-line AGN
seen along a lower inclination angle between the broad-line emitting region and the
line of sight (Baldi et al., 2016; Decarli et al., 2008; Rakshit et al., 2017).
Studies of NLSy1s typically find them to be highly photometrically variable
only in the X-rays. At optical wavelengths, however, Klimek et al. (2004) found that
rapid, high amplitude variability was rare in a sample of 172 observations of NLSy1s
across 33 nights. Ai et al. (2010) also found that NLSy1s had systematically lower
optical variability amplitudes (. 0.2 mag) than broad-line Seyfert 1s in a sample of
275 AGN at 0.3 < z < 0.8 in 3 years of SDSS data.
However, optical flares are not unheard of in NLSy1s (e.g. NGC 4051, Guainazzi
et al. 1998; Uttley et al. 1999). Klimek et al. (2004) noted the exception of IRAS
13224-3809, which showed both dramatic X-ray and optical variability on short
timescales (Miller et al., 2000). Here we describe a number of distinct events, in-
100
cluding the Trakhtenbrot et al. (2019a) observational class of optical flares, the
?on? state of AT2018dyk, and the host of PS16dtm and CSS100217:102913+404220,
which were all consistent with NLSy1 related activity.
Trakhtenbrot et al. (2019a) established a new observational class of dramatic
AGN flares accompanied by Bowen fluorescence features. The events in this class
all originated from active black holes that were classified as NLSy1 galaxies by their
Balmer FWHMs. Their optical spectra were unusual for NLSy1s in that they showed
strong ?double-peaked? He II profiles with contributions from the N III ?4640 Bowen
fluorescence feature, indicating the presence of a strong UV ionizing continuum. This
was consistent with the UV brightness observed in the small sample of objects as
well as the steep blue continua in these sources. The slow UV and spectral emis-
sion line evolution over a period of ?450 days ruled out a TDE, and these were
instead interpreted as enhanced accretion onto the SMBH of a pre-existing AGN.
AT2017bgt was presented as the prototype of these dramatic SMBH UV/optical
flares irradiating the BLR. It showed a very slow decrease in optical flux over sev-
eral months following a relatively shallow (?0.5 mag) rise to peak over ?80 days
from a previous non-variable state. During the transient, the X-rays increased by a
factor of 2? 3 from a previous measurement by ROSAT. The persistence of the UV
emission over 500 days distinguished it from SNe, and the extremely intense nature
of the UV continuum as well as presence of the Bowen fluorescence features in the
optical spectrum distinguished it from CLAGN. Two other NLSy1s, OGLE17aaj
(Gromadzki et al., 2019) and ULIRG F01004-2237 (Tadhunter et al., 2017) (the
latter previously interpreted as a TDE), were retroactively reclassified as belonging
to this new observational class of NLSy1s.
We compare with AT2018dyk, a changing-look AGN which transformed from
a LINER galaxy to a NLSy1. It was identified as such primarily based on X-ray and
UV spectra. It displayed strong high ionization forbidden (i.e. ?coronal?) emission
101
in the optical and UV spectra, an X-ray flare delayed by 60 days, and showed a
late-time g ? r color change as it faded slowly over 1 year. This was the only AGN
with Balmer lines consistent with a NLSy1 among a new class of ?changing-look
LINERs?, including SDSS 1115+0544 (Yan et al., 2019).
PS16dtm (iPTF16ezh/SN 2016ezh) was a near-Eddington but X-ray-quiet nu-
clear transient with strong Fe II emission and TBB ? 1.7? 104 K. It rose over ?50
days to ?superluminous? levels (log L ?1bol [ergs s ] > 44) at peak before plateauing
twice over ?50 and ?100 days while maintaining a constant blackbody temper-
ature. The event was interpreted as a TDE exciting the BLR in a well studied,
spectroscopically-confirmed NLSy1 with M 6BH ? 10 M(Blanchard et al., 2017). X-
ray upper limits showed dimming by at least an order of magnitude compared to
archival observations, but Blanchard et al. 2017 predicted the X-rays would reap-
pear after the obscuring debris (oriented perpendicularly to the accretion disk) had
dissipated. We show the V -band ASASSN photometry for PS16dtm in Figure 3.4
which appears similar in shape and absolute magnitude to AT2019fdr, though longer
in duration.
CSS100217:102913+404220 displayed a high state (MV = ?22.7 at 45 days
post-peak) accompanied by broad H? and was interpreted either as a Type IIn SN
(Drake et al., 2011) or a TDE (Saxton et al., 2018) near the nucleus (?150 pc) of a
NLSy1 in a star forming galaxy. It eventually faded back to slightly below its original
level after one year, which was interpreted as interacting with and subsequently
flushing a portion of the accretion disk.
Similar events are not unheard of in broad-line AGN systems, though they
may be comparatively more rare. Neustadt et al. (2020) reported a candidate for
such a rapidly flaring event with quasar-like properties, ASASSN 18jd, although
continued observations of this transient will be critical for a better understanding
of the properties of the host.
102
3.4.3 Observational Classification: The ?Family Tree? of NLSy1-
associated Transients
In Table 3.3, we use this sample to motivate a framework for quickly classifying
similarly ambiguous flaring events. We investigate the following:
? AGN/NLSy1 characteristics (an empirical W1?W2 WISE color cutoff from
Assef et al. 2013; Stern et al. 2012, which is comparable between NLSy1s and
broad-line Seyfert 1s; Chen et al. 2017; a strong Fe II complex; narrow Balmer
emission; and [O III]/H? < 3; Rakshit et al. 2017),
? TDE characteristics (host black hole mass below the Hills mass (? 108M),
and a lack of cooling or significant rebrightening),
? X-ray properties (the presence of which can occur in both AGN and TDEs,
but are less likely in the SN scenario).
We apply these criteria in Table 3.3 and color code them as blue or green based on
whether they favor the TDE or AGN scenario, respectively (as the SN scenario has
been ruled out in Section 3.4.1). The spectroscopic class, based on the presence of
N III Bowen fluorecence features, Fe II, and/or He II ?4686, which can occur in both
TDEs as well as flaring NLSy1s, is then interpreted in the context of one of these
scenarios. Based on this table, we confirm the interpretations for three of the four
NLSy1-associated transients reported in the literature, except for CSS100217 for
which we favor the AGN scenario over the SN interpretation.
Summarized briefly: We expect transients with strong Fe II complexes are most
likely associated with AGN, those with very steep soft X-ray spectra (? > 5) and
no intrinsic absorption are most likely associated with TDEs, and those with strong
Bowen fluorescence profiles and slow UV and spectral evolution are likely associated
with enhanced accretion onto supermassive black holes from a pre-existing accretion
103
disk. The timing of a mid infrared flare may also help to distinguish between an
AGN and a TDE ? if it precedes the optical, it is likely associated with AGN
variability, but if it follows as an echo, it may be associated with a TDE (van Velzen
et al., 2016).
van Velzen et al. (2020b) established a spectroscopic classification scheme for
the sample of TDEs discovered during the first half of the ZTF survey, distinguishing
those with and without He II in a single epoch. About half of the TDEs in that
sample were ?H-only?, and only one was ?He-only?. They found that higher density
conditions were likely for the rest of the TDEs which had H and He lines, as well as
Bowen features.
For the flaring NLSy1 sample presented here, we establish the following spec-
troscopic classes to describe each of the transients based on the presence or absence
emission features crucial to their physical interpretations:
1. ?He II only?,
2. ?He II+N III?, and
3. ?Fe II only?,
and we propose the following naming convention for these classes: ?NLSy1-HeII?,
?NLSy1-HeII+NIII?, and ?NLSy1-FeII?.10
3.4.4 Physical Interpretation of the Transient Flares
In the following section we consolidate all that is known about the relevant
properties of each object in the sample, and compare them with the related tran-
sients in NLSy1s in the literature, to explore each of the following scenarios: A
10We note that although hydrogen features are not explicitly named in this feature classification
scheme, all spectra of the transients show resolved narrow (1000 < FHWM < 2000 km s?1) Balmer
features (see Section 3.3.2).
104
Table 3.3: Comparison of the properties of individual objects in the sample (upper
table) and NLSy1-related transients in the literature (lower table). ?? means that
property is observed, and ??? indicates that characteristic was not observed. See
the extended published version in Frederick et al. 2021 (in prep.) for the full table
including UV and emission line diagnostics (uniform for all sources and representing
one for each class such that the interpretations in the final column are unchanged).
?Rebrighten? refers to a significant recovery of at least half the peak luminosity of
the source. Following the convention of Figure 3.7, blue (green) indicates a property
associated with the TDE (flaring AGN) scenario.
Name logMBH<8 H?<2000 Fe II ?g ? r X-ray ? W1-W2 Re- Spec. class Interp.
[M] km s?1 ? 0 mag >0.7 maga brighten
AT2019brs ?    b  ? HeII+NIII AGN
AT2019pev   ?  3 ?  HeII+NIII AGN
AT2019fdr    ? ? ? ? FeII TDE
AT2019avd    ? 5 ?  HeII+NIII AGN
AT2020hle   ?  b ? ? HeII TDE
CSS100217    ? 3  ? FeII AGN
PS16dtm     2c ? ? HeII+FeII TDE
AT2017bgt     2 ?  HeII+NIII AGN
AT2018dyk   ? ? 3 ? ? HeII AGN
PS1-10adi     b ? ? FeII TDE
a. We select the less conservative color cut presented in Stern et al. (2012).
b. The single low level XRT detection of AT2019brs and AT2020hle occurred only once throughout
the follow-up campaign and was not enough to take a reliable spectral measurement. Similarly, the
late stage X-ray detection of PS10adi reported in Jiang et al. (2019) was not sufficient to measure
the softness of the spectrum.
c. The host of PS16dtm displayed X-rays only prior to and following the fading of, but not for the
duration of, the transient.
PS16dtm-like TDE in a NLSy1, A Sharov-21-like microlensing event, a CSS100217-
like SN in a NLSy1, and a binary SMBH scenario.
3.4.4.1 Association of the Transients with AGN
There is evidence that all sources in the sample are associated with AGN rather
than distinct explosive events occurring in a normal galaxy. Although these out-
bursts may not necessarily be the result of an intrinsic enhancement in AGN accre-
tion activity, transients with fast-rise/slow-decay (such as those in this sample, along
with slow-rise/fast-decay, and symmetric light curve shapes) were well-represented
in a sample of 51 AGN flares discovered in CRTS (Graham et al., 2017).
105
Rakshit et al. (2017) spectroscopically classified the SDSS spectrum of the host
galaxy of AT2019brs as harboring an AGN NLSy1 > 12 years prior to the onset of
the smoothly flaring transient reported here.
As evident in Figure 3.9, the strengths of the Balmer lines in the transient
spectra are most consistent with that of a NLSy1. Ne V ?3426, when observable,
is typically associated with AGN, and is present in the spectra of these sources.
Strong He II profiles, although somewhat rare in association with normal stochastic
AGN variability (Neustadt et al., 2020), have been observed before and interpreted
as the signature of a sudden enhancement of accretion (e.g. Frederick et al. 2019;
Trakhtenbrot et al. 2019a).
Persistent X-rays are a likely signature of accretion onto a SMBH rather than
a SN. A strong soft X-ray excess is characteristic of NLSy1s. However, it is typically
accompanied by a hard X-ray continuum component (not present in either X-ray
detected transient in this sample), and not nearly as ultra-soft as the X-rays seen
in AT2019avd (4 . ? . 6), which are slopes more frequently observed in the X-ray
spectra of TDEs.
3.4.4.2 The SN Scenario
It is highly improbable that these flares are the result of normal SN explosions.
We observe long-lived U -band emission in AT2019fdr, persistent UV emission in all
transients in the sample, and strong transient X-ray detections in AT2019pev and
AT2019avd. There is also only a small likelihood of a SN in the host galaxy along the
line of sight unassociated with the AGN. The strongest evidence against the normal
supernova scenario is the persistence of the He II emission features ? 10? 100 days
after the onset of the flare ? such flash ionization signatures are only visible in
supernova spectra at very early times (e.g. Bruch et al. 2020; Khazov et al. 2016).
At least one of these transients (AT2019fdr) shares a number of properties with
106
CSS100217, which displayed soft X-rays and was interpreted as a SN IIn explosion
in an AGN disk. The SN interpretation of CSS100217 was largely based on light
curve energetics, which are similar to those of this sample. The g ? r color change,
and the peak magnitude of ?23 .MV . ?22 are very similar in particular between
CSS100217 and AT2019fdr. Type IIn supernovae can exhibit strong Fe II lines in
late spectra, such as AT2019fdr did.
However, in contrast, the light curve evolution differs in that CSS100217 fades
at least twice as quickly as AT2019fdr. Also, the Fe II complex of CSS100217 was
always visible throughout the flare, and Drake et al. (2011) observed a broad ?3000
km s?1 component in H? which got broader with time in subsequent follow-up
spectra of CSS100217. Strong P Cygni profiles are observed in the optical spectra of
SN, and from such profiles we would expect an absence of absorption on the blue end
of the Balmer line profiles, rather than emission as in the spectra of AT2019avd and
AT2019fdr. Therefore, based on this evidence we rule out the SN Type IIn scenario.
3.4.4.3 The TDE Scenario
The Hills mass is the mass for which the tidal Roche radius is equivalent
to the gravitational Schwarzschild radius of the black hole, beyond which a star
(that would otherwise be tidally pulled apart) is instead left intact as it passes the
event horizon (Hills, 1975). This maximum mass to tidally disrupt a solar-type
star just outside the event horizon is 108M. Therefore a SMBH mass significantly
above this limit would likely rule out a TDE. Of the supermassive black hole masses
derived for the host galaxies, only that of AT2019brs is inconsistent with a TDE
scenario, (although we note that it is consistent within the typical uncertainty for
such mass measurements). The range of absolute magnitudes of the flares in this
sample (?23 < Mr < ?19 mag) also tend to be intrinsically brighter at peak than
all but one of the ZTF TDEs (Mr > ?20 mag) reported in van Velzen et al. (2020a),
107
AT2018iih (Mr = ?21.5 mag).
Similar to TDEs PS16dtm and AT2018fyk, AT2019fdr showed two distinct
plateau stages on month-long timescales after fading, with some slight fading in
between. Color evolution is rare but not unheard of for TDEs, and the cooling
AT2019fdr shows post-peak is slow, with the transient still detected in the UV at late
times as would be expected for a TDE. Optical rebrightening following the intial flare
has been interpreted as the result of late time disk formation in a number of TDEs
(e.g. van Velzen et al. 2019; Wevers et al. 2019). However, rebrightening with high
amplitudes returning nearly to pre-flare levels such as that seen in AT2019pev and
AT2019avd has neither been observed11 nor predicted (e.g. Chan et al. 2020, 2019)
from a TDE. In these cases with rebrightening, a TDE is strongly ruled out.
AT2019avd and AT2019fdr, like AT2018fyk, only showed Fe II at certain
times during the flare. AT2019avd only displayed Fe II during its first peak,
and in AT2019fdr, the Fe II complex got more visible as the transient faded.
AT2019fdr is the only transient in the sample with a lack of He II features in its
spectra. Within the van Velzen et al. (2020b) spectral classification scheme for opti-
cal TDEs, AT2019fdr would be a H-only TDE, with the Fe II complex attributed to
the NLSy1 host. It is important to note that the transients with blue horn features
in H?, AT2019avd and AT2019fdr, may be signatures of wind ejecta with a velocity
distinct from the AGN.
Enhanced N III lines such as that seen in the NLSy1-HeII+NIII spectroscopic
class (AT2019pev, AT2019brs, and AT2019avd) are a prediction of TDEs in AGN
when compared to the host spectrum (Gallegos-Garcia et al., 2018; Kochanek, 2016;
Liu et al., 2018b). Unfortunately, a pre-flare spectrum was only available to test
this for AT2019brs (Figure 3.6).
Many properties of the hosts do not align with what we expect from AGN.
11Except in the case of the periodicity of ASASSN 14ko, which was interpreted as a possible
repeating partial TDE (Payne et al., 2020).
108
The WISE colors, for example, span a broad range of 0.06? 0.98 mag (Table 3.3).
The IR flare associated with AT2019avd could be interpreted as a dust echo, similar
to those seen in a number of TDEs (van Velzen et al., 2016). A host-subtracted SED
fit to the Swift photometry of AT2019avd gives a blackbody temperature consistent
with that of known TDEs, 104.25 K.
The X-ray variability of TDEs can vary erratically during a flare (e.g. van
Velzen et al. 2020b; Wevers et al. 2019). Although soft X-ray excesses with ? ? 3
are characteristic of NLSy1s, AT2019avd displays an X-ray power law index much
higher than typically seen in NLSy1s, and more characteristic of the extremely soft
X-ray spectra observed in TDEs.
Based on the combination of properties shown in Table3.3, we conclude that
two of the flares, AT2020hle and AT2019fdr, are better explained as TDEs than
AGN flares, although the interpretation is not clear-cut. However, if we assume
that their spectra are a combination of the host NLSy1 galaxy and the transient line
emission from the TDE, then given their spectral classes given here of NLSy1-FeII
and NLSy1-HeII, respectively, then the TDE spectra themselves, in these NLSy1
galaxies, would have to be of the TDE subclasses of TDE-H (H only lines) and
TDE-He (He II only lines), respectively, in order to match the observed spectra.
3.4.4.4 The Extreme AGN Variability Scenario
Graham et al. (2017) presented a sample of quasars displaying extreme vari-
ability in CRTS. Some had similar profiles and amplitudes (rising by 2?2.5 mag)
but longer timescales (500-1000 days) compared to the flares presented here, For
example, J002748-055559 rose by nearly ?2 mag compared to the steady level it
maintained for several years prior.
The optical spectra of the transients in the sample presented here belonging
to the ?NLSy1-HeII+NIII? spectroscopic class, as well as the UV brightness of the
109
sample, are consistent with the properties of the observational class of flares with
Bowen fluorescence established in Trakhtenbrot et al. (2019b). However, all of the
transients presented here have faster fading timescales than AT2017bgt. Trakht-
enbrot et al. (2019b) stated that the fade timescale of AT2017bgt was longer than
expected for a TDE. However, we note that at least one TDE in the van Velzen et al.
(2020b) sample (that also displayed Bowen fluorescence features) was observed to
fade over nearly 15 months,
3.4.4.5 The Gravitational Microlensing Scenario
Flares due to microlensing are expected to be observable in difference imaging
surveys with the combined baseline of iPTF and ZTF. The rise portions of the
light curve shapes of all the transients measured in Section 3.3.1.1 being well-fit by
quadratics is consistent with a lensing event, however, all but AT2020hle have a
longer decay with respect to the initial rise. Microlensing by multiple foreground
sources can give rise to a symmetric (with respect to the fade) double peak with a
dip in the middle of the optical light curve such as that seen in AT2019pev (Hawkins,
1998, 2004; Schmidt & Wambsganss, 2010). The cuspy shape of the first peak is
also characteristic of microlensing light curves. AT2019avd also showed a second
peak in its light curve, but the first peak was a lot more rapid and luminous than
the second. To test this scenario in AT2020hle would require continuing to observe
for an additional flare.
The microlensing scenario, however, would not account for the strong transient
Bowen fluorescence features that appear only at late times in AT2019avd, and only
at early times in AT2019pev (Figure 3.11). Meusinger et al. (2010) explained a
similar event as a background quasar with a UV flare in J004457+4123, also known
as Sharov 21, being microlensed by a foreground star in M31.
Microlensing is characteristically achromatic, and therefore would be ruled out
110
by the clear evidence for g ? r color change observed in AT2019fdr.
3.4.4.6 The SMBH Binary Scenario
Variability on the timescales of years due to a binary SMBHB system would
require a subparsec separation (e.g. Graham et al. 2015). In such a system, two
SMBHs induce tidal torques carving out a cavity in the circumbinary accretion
disk, and may be surrounded by their own minidisks at sufficient separations. The
interaction of accretion streams with the cavity could cause an outburst on the
approximate timescales seen in this sample, which is dependent on the properties of
the system. This phenomenon is seen in simulations of SMBH binaries (e.g. Gold
2019; Ryan & MacFadyen 2017).
We see evidence of offset narrow Balmer emission lines in the spectra of
AT2019fdr and AT2019avd, which may indicate a significant separate physical com-
ponent, although it is unclear what is contributing to those blueshifted velocities.
111
26 AT2019brs/ZTF19aailpwl (g) - 2.5 mag
24
22 AT2019fdr/ZTF19aatubsj (g) - 1 mag
20
AT2019pev/ZTF19abvgxrq (g) + 0.5 mag
18
AT2019avd/ZTF19aaiqmgl (g) + 1.5 mag
16
14 AT2020hle/ZTF18abjjkeo (g) + 5 mag
250 0 250 500 750 1000 1250
26
CSS100217 (V) - 2.2 mag
24
PS10adi (V) - 1.8 mag
22 PS16dtm (V)
20 AT2017bgt (V) + 1.2 mag
AT2018dyk/ZTF18aajupnt (g) - 2.5 mag
18
250 0 250 500 750 1000 1250
Days since rise to peak
Figure 3.3: The difference imaging light curves of the ZTF sample (upper panel)
compared to the published light curves of NLSy1-related events from the literature
(lower panel): changing-look LINER AT2018dyk (Frederick et al., 2019), TDE in a
NLSy1 PS16dtm (Blanchard et al., 2017), SN in a NLSy1 CSS100217 (Drake et al.,
2011), and the aperture photometry of flaring NLSy1 AT2017bgt (Trakhtenbrot
et al., 2019a). We show only g?band observations for the ZTF sample (upper
panel), and omit errorbars for visual purposes. Note the differences in optical filters
shown (g in green, V in blue), the differences in colors and markers used to represent
the same filters for visual clarity, as well as the difference in y-axis scale between the
panels. CRTS data for CSS100217 > 200 days prior to the transient is not shown
on this scale, but showed no significant activity for > 5 years.
112
Absolute Magnitude Absolute Magnitude
26
CSS100217
24 ilpwl
tubsj PS16dtm
22 bvgxrq
20 M =  -0.04 trise 18.59
bjjkeo
18 iqmgl ZTF18aajupnt
16
20 40 60 80 100 120
Rise to Peak (days)
Figure 3.4: Correlation of the rise times of the sample light curves with the maximum
absolute magnitude. Fits to light curves are described in Section 3.3.1.1. The same
color scheme and markers are used as in Figure 3.3.
113
Peak Absolute Magnitude
Velocity (km s 1) Velocity (km s 1) Velocity (km s 1)
4000 2000 0 2000 4000 4000 2000 0 2000 4000 4000 2000 0 2000 4000
1e 15 1e 16 1e 16
2.0 H ZTF19abvgxrq H ZTF19abvgxrq He II ZTF19abvgxrq
6
1.5 2
1.0 4
2 10.5
0.0 0 0
0.5 2
FWHM: 745 ? 47 km s 1 FWHM: 878 ? 49 km s 1 FWHM: 1183 ? 237 km s 1
1.0 4 1
6500 6550 6600 6650 4800 4825 4850 4875 4900 4925 4625 4650 4675 4700 4725 4750
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Velocity (km s 1) Velocity (km s 1) Velocity (km s 1)
4000 2000 0 2000 4000 4000 2000 0 2000 4000 4000 2000 0 2000 4000
1e 15 1e 16 1e 17
1.00 H ZTF19aatubsj 3 H l ZTF19aatubsj He II ZTF19aatubsj
0.75 6
2
0.50 4
0.25 1
2
0.00 0
0.25 0
1 1 1
0.50 FWHM: 448 ? 61 km s FWHM: 1208 ? 57 km s 2
6500 6550 6600 6650 4800 4825 4850 4875 4900 4925 4620 4640 4660 4680 4700 4720 4740 4760
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Velocity (km s 1) Velocity (km s 1) Velocity (km s 1)
4000 2000 0 2000 4000 4000 2000 0 2000 4000 5000 0 5000
1e 16
20 H ZTF19aailpwl H ZTF19aailpwl 2.0 He II ZTF19aailpwl
15 4 1.5
10 2 1.0
5 0.5
0 0 0.0
5 FWHM: 1603 ? 142 km s 1 2 FWHM: 1050 ? 77 km s 1 0.5 FWHM: 6325 ? 321 km s 1
10
6500 6550 6600 6650 4800 4825 4850 4875 4900 4925 4550 4600 4650 4700 4750 4800
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Velocity (km s 1) Velocity (km s 1) Velocity (km s 1)
4000 2000 0 2000 4000 4000 2000 0 2000 4000 7500 5000 2500 0 2500 5000 7500
1e 15 1e 16 1e 16
H ZTF19aaiqmgl H ZTF19aaiqmgl He II ZTF19aaiqmgl
1.5 4 3
1.0 2
2
0.5 1
0.0 0 0
0.5 1
FWHM: 1187 ? 24 km s 1 2 FWHM: 1433 ? 35 km s 1 FWHM: 6428 ? 1592 km s 1
1.0 2
6500 6550 6600 6650 4800 4825 4850 4875 4900 4925 4550 4600 4650 4700 4750 4800
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Velocity (km s 1) Velocity (km s 1) Velocity (km s 1)
4000 2000 0 2000 4000 4000 2000 0 2000 4000 7500 5000 2500 0 2500 5000 7500
1e 17
2.0 H ZTF18abjjkeo 0.4 H ZTF18abjjkeo 3 He II ZTF18abjjkeo
1.5 0.3 2
1.0 0.2
0.5 0.1 1
0.0 0.0 0
0.5 0.1
FWHM: 1289 ? 111 km s 1 0.2 FWHM: 1199 ? 270 km s
1 1 FWHM: 12 ? 620107 km s 1
1.0
6500 6550 6600 6650 4800 4825 4850 4875 4900 4925 4550 4600 4650 4700 4750 4800
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Figure 3.5: Gaussian fits to the H?+[N II] and H? line profiles of all transients
in the sample show that their Balmer lines have a FWHM consistent with (and
Lorentzian Balmer profiles characteristic of) that of narrow-line Seyfert 1s. The
offset blue peak in the H? profile of AT2019fdr is marked by a vertical line.
114
F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?)
F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?)
F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?)
F  (erg/cm2/s/?)
Mg II [OIII] [NeV] [O II] H? H? H? [O III] H? + [N II]
20
ZTF19abvgxrq (+0d)
ZTF19aailpwl (+34d)
15 ZTF19aailpwl (-13y)
ZTF19aatubsj (+368d)
ZTF?19aaiqmgl (+444d)
10
ZTF18abjjkeo (+8d)
AT 2017bgt (Trakhtenbrot+2019)
?
?PS16dtm (Blanchard+2017)
5
CSS100217 (Drake+2011)
CSS100217 (Abazajian+2009)
0 Fe II ZTF18aajupnt (Frederick+2019)
3000 4000 5000 6000 7000 8000
Rest Wavelength (A?)
Figure 3.6: Comparison of the ZTF sample of flares (in blue), as well as discovery
spectra for the NLSy1-related events from the literature (in black): changing-look
LINER AT2018dyk (Frederick et al., 2019), TDE in a NLSy1 PS16dtm (Blanchard
et al., 2017), SN in a NLSy1 CSS100217 (Drake et al., 2011), and Bowen fluorescent
flare AT2017bgt (Trakhtenbrot et al., 2019a), and their pre-event spectra when
available (in grey). For AT2019fdr and AT2019avd here and in Figure 3.7, we plot
the spectra after continuum fading rather than the discovery spectra, to display the
features used in the spectroscopic classification scheme discussed in Section 3.4.3.
AT2019fdr and AT2019avd show offset blue peaks in H?, and the peak of He II is
offset from 4686 A? in AT2020hle.
115
Normalized F? + Constant
He II
He II
[O II] H?/[N III] H?/[O III] H? [OIII]
Fe II only
He II only
17.5 He II+Bowen Fe II
15.0
CSS100217 (Drake+2011)
12.5
ZTF19aatubsj (+368d)
ZTF18aajupnt (Frederick+2019)
10.0
PS16dtm (Blanchard+2017)
7.5
ZTF18abjjkeo (+8d)
ZTF19aailpwl (+34d)
5.0
ZTF19abvgxrq (+0d)
2.5
ZTF19aaiqmgl (+444d)
AT 2017bgt (Trakhtenbrot+2019)
0.0
3800 4000 4200 4400 4600 4800 5000
Rest Wavelength (A?)
Figure 3.7: Zoom-in on the 4000-5000 A? region of Figure 3.6 showing the comparison
of the strength of H?, Fe II, and [O III] of the sample with NLSy1-related events
in the literature. We color-code the sample and establish categories based on the
presence and absence of key emission line features as described in Section 3.4.3.
Blue spectra indicate the presence of He II, and black spectra indicate transients
which displayed Fe II only, (though we note that PS16dtm showed both features).
Those in green display Bowen fluorescence features in addition to He II, and are
most spectroscopically similar to AT2017bgt and the other AGN flares comprising
the sample in Trakhtenbrot et al. (2019a).
116
Normalized F? + Constant
O III
N III
He II
power law (? = 2.7? 0.1)
Background
ZTF19abvgxrq XRT stacked
10?3
10?4
10?5
10?6
100 101
2
1
0
100 101
Energy (keV)
power law (? = 5.7? 0.5)
Background
ZTF19aaigmql XRT stacked
10?3
10?4
10?5
10?6
15 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
10
5
0
0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Energy (keV)
Figure 3.8: Upper panel: An absorbed power law fit and ratio residuals to the ?100
ks stacked Swift XRT spectrum of AT2019pev (spectral index ? = 2.7?0.1). Lower
panel: The ?4 ks stacked spectrum of AT2019avd (? = 5.7? 0.5).
117
normalized counts s?1 keV?1 cm?2 normalized counts s?1 keV?1 cm?2
Ratio [Data/Model] Ratio [Data/Model]
[OIII] [NeV] [O II] H? H? H? [O III] H? + [N II]
AT2019brs/ZTF19aailpwl +34d
20 AT2019pev/ZTF19abvgxrq +0d
AT2019fdr/ZTF19aatubsj +368d
15
AT2019avd/ZTF19aaiqmgl +444d
AT2020hle/ZTF18abjjkeo +8d
10
NLS1 Mrk 618
5 SN IIn 2005bx -15d (Kiewe+2012)
TDE AT2019dsg (van Velzen+2020)
0
3000 4000 5000 6000 7000 8000
Rest Wavelength (A?)
Figure 3.9: We compare the spectra of the transient sample (in black) to archetypal
NLSy1 Mrk 618, as well as a normal Type IIn supernova, SN 2005bx (Kiewe et al.,
2012), and AT2019dsg, a normal TDE in a star forming galaxy with Bowen fluo-
rescence features and a coincident neutrino detection (Stein et al., 2020; van Velzen
et al., 2020b).
118
Normalized F? + Constant
He II
He II
3.5 Conclusions
We report five nuclear flaring events associated with NLSy1s, all serendipi-
tously12 discovered in ZTF. We measured their photometric characteristics (such
as light curve shape, g ? r color, and rise to peak luminosity, finding a correla-
tion between rise time and absolute magnitude), and spectroscopic properties. We
then established groupings of the objects in the sample based on analyses of the
months-long follow-up campaigns of these objects. Based on observed groupings of
the sample, we propose the following naming scheme of spectroscopic classes of such
transients for use in future optical surveys: ?NLSy1-HeII?, ?NLSy1-HeII+NIII?,
and ?NLSy1-FeII?. We ruled out the possibility that these are Type IIn supernovae
occurring in NLSy1 systems. Despite the heterogeneity of the sample?s properties,
two of the flares presented in this work have multiwavelength characteristics which
could be consistent with TDEs in NLSy1s (AT2019fdr and AT2020hle), with spec-
tral classes of NLSy1-FeII and NLSy1-HeII, respectively. This is a high TDE rate
relative to quiescent galaxies, which are more abundant than NLSy1s. The preva-
lence of TDE candidates in the NLSy1 AGN class could be a natural result of their
hosting smaller black holes compared to typical broad-line AGN, and therefore sat-
isfying the Hills mass criterion for an observable TDE. However, without pre-event
spectra and X-ray imaging to isolate the contribution of the putative TDE to the
composite NLSy1+TDE emission, flaring due to extreme AGN variability cannot be
definitively ruled out. For two in the sample (AT2019pev and AT2019avd), we can
rule out the simple TDE scenario from rebrightening in their light curves, and we
determine that they, along with AT2019brs (which had a pre-flare NLSy1 spectral
classification and a black hole mass estimate too large to host a canonical TDE), are
12As the Trakhtenbrot et al. (2019a) observational class was established midway through the
ZTF survey, we had not been systematically filtering such events when the population became
apparent in the nuclear transients alert stream search.
119
likely outbursts related to enhanced accretion in excess of typical AGN variability,
and with spectral features we classify as ?NLSy1-HeII+NIII?, and members of the
Trakhtenbrot et al. (2019a) class of AGN flares.
Given this sample, together with the growing number of interesting rapid
optical transients associated with NLSy1s we reviewed in the literature, we posed
the question of why such environments are observed to preferentially host these
outbursts. Given the relative fraction of NLSy1s found with respect to other AGN
classes in spectroscopic surveys such as SDSS (?15%; e.g. Rakshit et al. 2017; Zhou
et al. 2006), there is likely an underlying factor enhancing this rate. We suggest
four different possible explanations for this enhancement:
1. A selection bias due to shorter timescales for lower mass BH systems (like
NLSy1s), which are therefore more likely to be captured within the baseline
of wide field optical surveys,
2. A systematic disregard of smooth flares in broad line AGN during transient
searches, or
3. A true intrinsic rate enhancement due to instabilities causing rapid changes
in the observable environments or accretion efficiencies of these systems.
Follow-up strategies of optical transients in AGN that are similarly ambiguous at
early times may stand to benefit from the framework we offer here. We hope this
classification scheme will guide real-time predictions for potential future behavior
of large amplitude flares in NLSy1s, which are clearly an interesting population for
future study. The next step will be to perform a systematic study of the variability
of NLSy1s detected in ZTF, to assess the completeness and rate of this sample of
transients with smoothly flaring light curves, and compare to a sample of broad-line
AGN. Expanding on the small number of unusual transients associated with NLSy1s
120
not only sheds light on the parameter space in which they reside, but also provides
the framework for a decision tree for understanding such outbursts when they are
inevitably captured at higher rates in upcoming wide field surveys. This will be
imperative to establish in advance of larger and deeper surveys such as ZTF Phase
II and the Vera C. Rubin Observatory (formerly known as LSST; Ivezic? et al. 2019),
to which the timescales of these flares are well-suited. Continued multiwavelength
monitoring of the entire sample will be important to determine the host properties
for those with sparse data prior to the transient, and for understanding the evolution
and nature of these flares.
3.6 Appendix
The light curves in Figure 3.10 are from the second IPAC data release of
ZTF forced photometry. Figure 3.11 shows the region of interest around He II,
H?+[O III], and the Fe II complex for all follow-up spectra taken of the sample.
121
18 17
18
19
19
20
20
21 21
22 22
0 200 400 600 ?400 ?200 0
Days since discovery Days since discovery
AT2019brs AT2019pev
18 18
20 20
22 22
?400 ?200 0 200 400 ?200 ?100 0 100 200 300 400
Days since discovery Days since discovery
AT2019fdr AT2019avd
18
19
20
21
22
?200 0 200 400
Days since discovery
AT2020hle
Figure 3.10: Forced photometry of the sample from ZTF Data Release 3. Colors
correspond to r-, g-, and i-band 3-? detections, and triangles correspond to 5-?
upper limits. An ?X? marks the rise to peak in the difference imaging light curve
of AT2020hle, which was discovered in data too recent to be included in the ZTF
DR3, and therefore only shows the flux level of the host galaxy. The data points
in the light curves beyond 2020 will be released in the final ZTF photometry data
release.
122
Forced Photometry Magnitude Forced Photometry Magnitude
Forced Photometry Magnitude
Forced Photometry Magnitude Forced Photometry Magnitude
[O II] H?/[N III] H?/[O III] H? [OIII] [O II] H?/[N III] H?/[O III] H? [OIII]
Fe II Fe II
25
8
ZTF19aailpwl +666d LDT ZTF19abvgxrq +354d LDT
20 ZTF19abvgxrq +154d LDT
ZTF19aailpwl +413d P60
6 ZTF19abvgxrq +69d LDT
15 ZTF19abvgxrq +28d LDT
ZTF19aailpwl +133d LDT
4 ZTF19abvgxrq +17d LT
ZTF19aailpwl +34d FLOYDS-N 10 ZTF19abvgxrq +10d Lick
ZTF19abvgxrq Keck1
2
ZTF19aailpwl -4605d SDSS 5 ZTF19abvgxrq LDT
0 0
3800 4000 4200 4400 4600 4800 5000 3800 4000 4200 4400 4600 4800 5000
Rest Wavelength (A?) Rest Wavelength (A?)
[O II] H?/[N III] H?/[O III] H? [OIII] [O II] H?/[N III] H?/[O III] H? [OIII]
Fe II Fe II
ZTF19aatubsj +506d LDT
17.5 ZTF19aaiqmgl +579d P60
10
ZTF19aatubsj +408d LDT
15.0
ZTF19aatubsj +403d P60 ZTF19aaiqmgl +576d P60
8
12.5 ZTF19aatubsj +368d NOT
ZTF19aaiqmgl +575d P60
10.0 ZTF19aatubsj +104d P60 6
ZTF19aaiqmgl +444d FLOYDS-S
ZTF19aatubsj +66d P200
7.5
4
ZTF19aatubsj +55d LDT ZTF19aaiqmgl +443d P60
5.0
ZTF19aatubsj +50d LT
2 ZTF19aaiqmgl +22d NOT
2.5 ZTF19aatubsj +45d P60
0.0 0
3800 4000 4200 4400 4600 4800 5000 3800 4000 4200 4400 4600 4800 5000
Rest Wavelength (A?) Rest Wavelength (A?)
[O II] H?/[N III] H?/[O III] H? [OIII]
9
Fe II
8
7
6
5 ZTF18abjjkeo +210d LDT
4
ZTF18abjjkeo +8d LT
3
2
ZTF18abjjkeo +6d P60
1
0
3800 4000 4200 4400 4600 4800 5000
Rest Wavelength (A?)
Figure 3.11: Spectroscopic follow-up of the sample summarized in Table 3.1, showing
the evolution of the He II, H?, and Fe II line complexes.
123
Normalized F? + Constant Normalized F? + Constant Normalized F? + Constant
O III O III
O III
N III N III
He II N IIIHe II He II
Normalized F? + Constant Normalized F? + Constant
O III O III
N III N III
He II He II
Chapter 4: The X-ray View of a New Class of Changing-look LIN-
ERs
ZTF has enabled the discovery of a new class of LINERs ?turning on? into
AGN with dramatic optical spectroscopic transformations. For the most rapid tran-
sition into a narrow-line Seyfert 1, real-time monitoring revealed the presence of a
prominent soft excess and a luminous X-ray flare delayed with respect to the optical
rise by 2 months. We present new Swift, XMM-Newton, and NuSTAR follow-up
observations of this sample, as well as optical spectroscopic follow-up data, and in-
troduce additional sources discovered in the second half of the 3-year ZTF Phase I
survey. We also view for the first time this new class of changing-look LINERs in
the hard X-rays, and contrast their optical, UV, and X-ray properties with that of
broad-line and changing-look Seyferts. We observe a change in H? and [O III] line
ratios over several years of follow-up observations of the sample, and observe the
sample in various stages of X-ray hardening with 1.0 < ?OX < 1.6. The X-ray data
in particular point to an intriguing connection between the accretion modes of this
observational class of changing-look AGN and the strength of the observed X-ray
spectral components, specifically a hardening over time.
124
4.1 Introduction
?Changing-look? active galactic nuclei (CLAGN, also referred to as ?changing-
state? AGN)1 are a growing class of objects that are a challenge to the simple AGN
unification picture. They demonstrate the appearance (or disappearance) of broad
emission lines and a non-stellar continuum, changing their classification between
narrow-line and broad-line AGN on a timescale of months to years. Systematic stud-
ies at various wavelengths show that the nature of these spectral transformations
is predominantly driven by changes in accretion rate, but the mechanism or mech-
anisms driving these sudden changes is still not well understood (LaMassa et al.,
2015; MacLeod et al., 2016, 2019; Ruan et al., 2016; Runnoe et al., 2016). Evidence
in recent years points to changing-state AGN including heterogenous observational
sub-classes of objects (e.g. Mg II CLQs; Guo et al. 2019), with some occupying a
higher echelon of the continuous distribution of stochastic AGN variability, and oth-
ers representing more extreme and exotic accretion events than previously observed
(e.g. Frederick et al. 2019).
During the first year of the Zwicky Transient Facility Survey, a new class of
dramatic changing look AGN (CLAGN) was uncovered to occur in LINER galaxies
(Frederick et al., 2019). One of these (ZTF18aajupnt/AT2018dyk) showed soft X-
ray flaring delayed by 2 months with respect to the optical rise. As this was the only
object in the sample to be promptly followed up with multiwavelength resources, it
is therefore largely unknown what their X-ray properties are as a class.
Parker et al. (2016) show that X-ray observations of CLQs can be a powerful
tool in understanding the physical mechanisms of an optical changing look (distinct
from X-ray changing-look AGN, which show variability in hydrogen column densi-
ties) by measuring the response of the reprocessed hard X-ray continuum. Coupled
1Optical CLAGN are distinct from X-ray AGN ?changing-look? between Compton-thick/-thin
class.
125
with mapping out the structure of the accretion flow state change in optical spec-
tra, a more complete story can be told about what is occurring in each region of
an AGN during its state transition. Ruan et al. (2019a) showed a direct analogy
of samples of changing-look quasars with the well-studied state transitions X-ray
binaries (XRBs). In XRBs, the ratio of the X-ray and UV flux densities probe the
disk-corona relationship as the system passes through cyclical states characterized
by accretion rate. In the low/hard state, the coronal emission dominates over the
dimmer UV emission from a truncated disk caused by a jet or advection dominated
accretion flow (ADAF), and in the high/soft state, the UV emission from the disk
dominates the spectrum. Ruan et al. (2019a) scaled state changes of X-ray bina-
ries to changing-look AGN in various stages of accretion and showed similar state
changes in both according to predicted ?OX and Eddington ratios. Ruan et al.
(2019b) showed the evolution of the state change in individual CLAGN (including
AT2018dyk) followed a well-studied XRB along this parameter space. They mea-
sure softening below a critical Eddington ratio threshold of ? 1% explained as disk
reprocessing of Comptonized X-rays from the corona. Jin et al. (2021) found fur-
ther agreement with this work by comparing Eddington ratios and flux densities
from multi-epoch X-ray and UV/optical snapshots for a sample of 10 known CLQs
in various stages of evolution to simulated XRB data. They also uncovered a link
between dramatic changes in the strength of optical emission lines and the evolution
of objects across this inversion point.
To better understand the evolution of both the optical and X-ray properties
of the sample of CL-LINERs, we undertook a 3-year follow-up campaign consist-
ing of optical and X-ray spectroscopic observations with XMM-Newton, NuSTAR,
Swift, and various ground based facilities. We sought to detect any evolution of
both the broad and narrow lines in the optical spectra as well as the continuum,
and the presence of components signaling reflection in the X-ray spectra. Coronal
126
X-rays irradiating the accretion disk induce fluorescence, the most prominent emis-
sion being from Fe (George & Fabian, 1991). A number of features arise from the
resulting reflection spectrum, due to reflection at all distances along the accretion
disk, and elucidate the interplay between absorption, line emission, and scattering
at these high energies (Fabian & Ross, 2010). One of these features, the ?Compton
hump?, is evident in X-ray spectra above 10 keV and interpreted as reflection of
> 80 keV continuum photons by the accretion disk or as far out as the molecular
torus via Compton down-scattering. Another prominent feature is the iron line, and
when relativistically broadened it can be a robust measurement of how emission is
reprocessed in the strong gravity region near the central black hole. The presence
or absence of these X-ray spectral features following a changing-look transition pro-
vides an alternate method for mapping the environment of the accretion flow in
these sources, contributing to an understanding of the nature of the related physical
processes, and exhibiting how they compare to normal AGN.
We present these X-ray and optical follow-up observations of this new class of
CL-LINERs, and introduce additional sources discovered in the second half of the
3-year ZTF Phase I survey.
We assume the following cosmology for calculations relying on distance mea-
surements: H = 70 km s?1 Mpc?10 , ??= 0.73 and ?M = 0.27. RA and Dec are
reported at J2000.0.
4.2 Observations
All follow-up observations for this sample are summarized in Table 4.7 of the
Appendix.
127
4.2.1 Target Selection
For this study we largely follow the sample selection criteria and multiwave-
length data reduction procedures outlined in Frederick et al. (2019). CL-LINER
candidates were triggered for follow-up on the basis of their variable ZTF photom-
etry and cross-matching with galaxies classified as LINERs and displaying Balmer
emission in the Portsmouth Sloan Digital Sky Survey (SDSS) DR12 catalog (Thomas
et al., 2013). The CL LINERs were selected for follow-up on the basis of having con-
secutive detections in ZTF difference imaging, with variability above the threshold
of photometric error (&0.1 mag).
4.2.2 Archival Data
The host galaxies of this sample all had Balmer and low ionization forbidden
line ratios consistent with that of LINER galaxies, as published in the Portsmouth
catalog. The ?off? state of ZTF19aambzmf was additionally classified as a Sy1.9
type galaxy in the Ve?ron-Cetty et al. (2001) catalog, due to having a broad base in
the profile of the H? emission line in the 2007 SDSS spectrum. The optical archival
observations of the sample were presented in Frederick et al. (2019).
To compare with the new X-ray follow-up data, we summarize the available
X-ray archival observations of the sources presented in this work, from NASA High
Energy Astrophysics Science Archive Research Center (HEASARC), the XMM-
Newton science archive, and the UK Swift Science Data Centre. The detections
as well as most recent 3-? X-ray upper limits2 of the sample are provided in Ta-
ble 4.1. For these upper limit calculations, Galactic column densities were estimated
as N = 1 ? 1020 cm?2H for all sources except for ZTF18aaabltn, which is approx-
imated as NH = 3 ? 1020 cm?2, following the HI4PI Collaboration et al. (2016)
2http://xmmuls.esac.esa.int/upperlimitserver/
128
Table 4.1: Archival ROSAT, Swift XRT, and XMM-Newton Slew Survey 3-? De-
tections and Upper Limits of CL LINER candidates
Source z Instrument Observation Exposure Energy Count FX LX
Name Name Date Time Range Rate
(UTC) (s) (keV) (counts s?1) (10?12) (1043)
ZTF18aajupnt 0.0367 ROSAT 1990 Jul 30 742.9 0.2?2 < 0.0347 < 0.2452 < 0.1
XMM-Newton 2014 Aug 17 4.0 0.2?12 < 2.2 < 4.6 < 1.5
ZTF18aahiqfi 0.067 ROSAT 1990 Nov 19 470.4 0.2?2 < 0.03 < 0.2 < 0.3
XMM-Newton 2003 Dec 26 8.7 0.2?12 < 1.2 < 2.5 < 2.8
ZTF18aaidlyq 0.1 ROSAT 1990 Oct 10 398.5 0.2?2 < 0.05 < 0.4 < 1.0
XMM-Newton slew 2011 May 05 6.6 0.2?12 < 1.2 < 2.5 < 6.4
ZTF18aasszwr 0.168 ROSAT 1990 Nov 21 399.9 0.2?2 0.03 ? 0.01 0.19? 0.07 1.5?0.6
XMM-Newton 2014 May 15 4.5 0.2?12 1.7915 ? 0.7375 3.6941? 1.5209 30?10
ZTF18aaabltn 0.045841 ROSAT 1990 Oct 13 327.3 0.2?2 < 0.07 < 0.7 < 0.3
XMM-Newton 2015 Apr 08 9.3 0.2?12 0.8 ? 0.3 1.7? 0.8 0.9?0.4
ZTF18aahmkac 0.0699 ROSAT 1990 Nov 30 436.8 0.2?2 < 0.05 < 0.3 < 0.4
Swift-XRT 2012 Apr 07 942.7 0.2?12 0.0044?0.0037 0.17? 0.14 0.21?0.17
ZTF18aaavffc 0.0654 ROSAT 1990 Dec 05 504.8 0.2?2 0.14?0.02 1.0? 0.1 1.0?0.1
XMM-Newton 2018 Dec 30 6.9 0.2?2 < 2.0 < 2.5 < 2.7
ZTF19aambzmf 0.122 ROSAT 1991 Jan 03 387.0 0.2?2 < 0.06 < 0.4 < 1.7
measurements presented in Section 4.4.1. The count rate to flux conversion calcu-
lation was performed assuming an X-ray power law spectral index of ?=2.
Similar to the photometric analysis of the sample presented in Frederick et al.
(2019), we investigate the archival Catalina Real-time Transient Survey (CRTS)
V -band light curves and All-Sky Automated Survey for Supernovae (ASASSN3;
Jayasinghe et al. 2019; Kochanek et al. 2017; Shappee et al. 2014) aperture pho-
tometry of the newest objects in this CL-LINER sample in order to constrain the
transition timescales prior to their discovery in ZTF difference imaging. This ?t
listed in Table 4.2 is defined from the gradual increase in the optical variability to
the confirmation of the classification change in spectroscopic follow-up data.
ZTF18aaavffc ? The CRTS data of this source has an upward trend but shows
no variability from a baseline of V = 15.7 mag above the 0.2 mag level prior to 2013,
6 years before its discovery in ZTF. It was detected in ASASSN starting in Feb 2012,
and showed a slow gradual rise from 16.0 to 15.5 mag in g-band difference imaging
photometry starting around 2017.
ZTF19aambzmf ? This source was detected steadily at V = 17.1 ? 0.2 mag
3https://asas-sn.osu.edu/
129
in CRTS between 2005 and 2013, 6 years before its discovery in ZTF. This source
was also detected in ASASSN difference imaging at a similar level starting in 2012.
ZTF18aahmkac ? The sparse CRTS data of this source shows no variability
from V = 16.35 mag above the 0.1 mag level prior to 2012, 6 years before its
discovery in ZTF.
4.3 Optical
The majority (5 out of 9) this sample were discovered in ZTF difference imag-
ing for the first time in April and May of 2018, soon after ZTF Phase I Science
Commissioning commenced. This indicates these sources were likely variable prior
to their detections in ZTF, and is therefore not a representative sample selection.
This was confirmed for the subset of this sample discovered in year 1 of ZTF Phase
I presented in Frederick et al. (2019), which showed the slow onset of the optical
variability in forced and PSF photometry from ZTF and other transient surveys,
which define the optical transition timescales presented in that study as being earlier,
sometimes by several years, than the discovery in ZTF difference imaging. These
transition timescales, as well as other properties of the host galaxies and changing-
look states, are listed in Table 4.2 for the sources in the original sample as well
as 3 additional sources discovered in the final half of ZTF Phase I (from mid 2019
through Oct 2020). iPTF 16bco (Gezari et al., 2017), which is included in this
sample as the first CL-LINER, was observed with ZTF in 2018 (2 years after its
discovery in iPTF and ?1 year after detection by the XMM Slew Survey), and is
also known by its ZTF name as ZTF18aajarpg.
130
Table 4.2: Summary of the optical properties of the changing-look LINER sample
from the Zwicky Transient Facility (ZTF) Survey Phase I, adapted and extended
from Frederick et al. (2019), with sources new to this sample listed in the top panel.
We list redshifts from the Portsmouth SDSS DR12 catalog (Thomas et al., 2013).
Transition timescales ?t are roughly constrained based on the time delay between
the onset of variability detected in the host in the archival light curves, and the time
of the first spectrum taken in the type 1 AGN state. Estimates of star formation
rate by Chang et al. (2015) are from SDSS+WISE SED model fitting. ?m is the
variability magnitude change with respect to the host galaxy magnitude defined in
Eq. 3 of Hung et al. (2018).
Name RA Dec z DLum Discovery Date MDiscovery ?t Host Type
4 log SFR ?mvar High State
(hh:mm:ss.ss) (dd:mm:ss.ss) (Mpc) (mag) (yr) [M yr?1 ] (mag)
ZTF18aahmkac 11:58:18.02 +10:03:22.6 0.0699 310 2018 Apr 8 ?17.82 <0.4 Spiral 0.162 ?0.06 quasar
ZTF18aaavffc 12:31:55.15 +32:32:40.3 0.0654 290 2019 Oct 23 ?18.37 <0.5 Elliptical 0.907 ?0.27 quasar
ZTF19aambzmf 14:40:21.48 +14:11:25.9 0.122 570 2019 Mar 1 ?18.53 <1.0 Elliptical 0.232 ?0.10 NLSy1
ZTF18aajupnt 15:33:08.01 +44:32:08.2 0.0367 158 2018 May 31 ?16.59 <0.3 Spiral 0.177 ?0.18 NLSy1
ZTF18aasuray 11:33:55.83 +67:01:08.0 0.0397 171 2018 May 10 ?17.80 <6.8 Spiral 0.147 ?0.06 Seyfert 1
ZTF18aahiqfi 12:54:03.80 +49:14:52.9 0.0670 296 2018 April 8 ?18.25 <0.6 Elliptical ?0.058 ?0.12 quasar
ZTF18aaidlyq 09:15:31.06 +48:14:08.0 0.1005 457 2018 April 11 ?19.09 <0.7 Spiral 0.092 ?0.29 quasar
ZTF18aaabltn 08:17:26.42 +10:12:10.1 0.0458 199 2018 Sept 15 ?17.62 <2.6 Elliptical 0.227 ?0.81 quasar
ZTF18aasszwr 12:25:50.31 +51:08:46.5 0.1680 813 2018 Nov 1 ?20.40 <5.3 Elliptical 1.267 ?0.72 quasar
iPTF16bco5 15:54:40.26 +36:29:52.4 0.2368 1197 2016 June 1 ?20.76 <1.1 Elliptical 1.007 ?0.69 quasar
4.3.1 ZTF Photometry
We show the heterogeneous set of ZTF g and r band photometry for this
sample in Figure 4.1, beginning with their discovery which requires a 5-? signif-
icance detection above the reference image of the LINER host galaxy. The ab-
solute magnitudes at the time of discovery are given in Table 4.2. The objects
display a large range in absolute magnitudes, with minimum and maximum val-
ues of M between -16 and -21. Some, like ZTF18aahiqfi, ZTF18aaabltn, and
ZTF19aambzmf show stochastic variability in the ZTF difference imaging photom-
etry, while ZTF18aasszwr, AT2018dyk, ZTF18aasuray, and ZTF18aaavffc show in
addition to this some smoother, larger amplitude flare features.
In Figure 4.2 we also show the ZTF forced photometry provided by the NASA/IPAC
Infrared Science Archive6 for the latest discovered sources ZTF18aahmkac, ZTF19aambzmf,
and ZTF18aaavffc. These light curves show the timing of the rise to peak not seen
in difference imaging, which requires a 5-? detection above the level of the reference
6https://irsa.ipac.caltech.edu/Missions/ztf.html
131
s s x sx ZTF18aaidlyq
?19
?18
s x s x ZTF18aasszwr
?21
?20
?19
s x ZTF18aahiqfi
?18
?17
s xs x ZTF18aaabltn
?19
?18
?17
s sssus u u s s x x ZTF18aajupnt
?17
?16
s x s ZTF18aahmkac
?18
?17
s x s ZTF18aasuray
?18
?16
ss x ZTF18aaavffc
?20.0
?17.5
s x ZTF19aambzmf
?19
?18
0 200 400 600 800 1000 1200 1400
Days Since Discovery
Figure 4.1: Comparison of the ZTF g- and r-band difference imaging light curve
shapes of the sample. The timing of the optical spectroscopic follow-up epochs with
respect to the ZTF photometry in absolute magnitude are indicated alongside each
light curve with an ?s?, X-ray follow-up with an ?x? and Swift UVOT follow-up with
a ?u?. All follow-up data are summarized in detail in Table 4.7 of the Appendix.
132
Absolute Magnitude
image and therefore does not capture the entirety of the optical transition to a larger
amplitude variability state. See Figure 2 of Frederick et al. (2019) for the remainder
of the sample.
16.5
17.0 ?
17.5 ZTF18aahmkac
? ZTF19aambzmf
18.0 ZTF18aaavffc
18.5 ?
19.0
58200 58400 58600 58800 59000 59200
MJD
Figure 4.2: The ZTF g-band forced photometry for the three new objects in this
sample introduced in this work. The rise time for each source corresponding to the
changing-look transition timescale ?t measured in Table 4.2 is marked with ???.
4.3.1.1 Optical Monitor
We use the SAS command omichain to reduce the UVW1 (effective wavelength
2910 A?) XMM Optical Monitor data. The Vega magnitudes for the sample ranged
widely from 12 to 20 mag. See Section 4.3.3.2 for more details about the XMM
observations.
133
Apparent Magnitude
4.3.2 Spectroscopy
The data from the follow-up campaign for this sample are shown in Figures 4.11
and 4.3, and summarized in the Appendix in Table 4.7. This set of follow-up data
included spectroscopy taken by the robotic SED Machine mounted on the Palomar
60-inch Observatory (SEDM/P60; Program PIs: Gezari, Sollerman, Kulkarni), the
Hale Telescope at the Palomar 200-inch Observatory (PI: Yan), the Deveny Spectro-
graph at the Lowell Discovery Telescope (PI: Gezari), and GMOS-North at Gemini
Observatory (PIs: Hung, Frederick). We reduced P60 spectra with pysedm (Rigault
et al., 2019), and all other spectra with pyraf using standard procedures.
Compared to these archival spectra, all sources in the sample showed dra-
matic changes in their broad emission line profiles, often with bright blue con-
tinua. These spectral transformations for the three latest CL-LINERs added to
the sample (ZTF18aahmkac, ZTF18aaavffc, ZTF19aambzmf) are shown in Fig-
ure 4.3. ZTF18aahmkac shows broad double peaked emission in H? and H? in ad-
dition to a strong blue continuum emergent from a LINER spectrum 15 years prior.
ZTF18aaavffc shows a broad component in H? only in the archival spectrum and
the Balmer emission lines appear broader in both follow up spectra. ZTF19aambzmf
shows broader Balmer emission compared to the archival spectrum from 12 years
prior, but relatively less change in the continuum compared to the other CL-LINERs
in the sample. Follow up spectral epochs for the rest of the sample are described in
detail in Section 4.4.2.
134
[O II] H? H? H? [O III] H? + [N II] [O II] H? H? H? [O III] H? + [N II]
7
4.0
ZTF18aahmkac +415d LDT
3.5 6 ZTF18aaavffc +40d LDT
3.0 5
2.5 ZTF18aaavffc +31d P60
4
2.0
3
ZTF18aahmkac -5483d SDSS
1.5
2 ZTF18aaavffc -5307d SDSS
1.0
1
0.5
0.0 0
4000 4500 5000 5500 6000 6500 7000 4000 4500 5000 5500 6000 6500 7000
Rest Wavelength (A?) Rest Wavelength (A?)
[O II] H? H? H? [O III] H? + [N II]
4
ZTF19aambzmf +178d LDT
3
2
ZTF19aambzmf -4309d SDSS
1
0
4000 4500 5000 5500 6000 6500 7000
Rest Wavelength (A?)
Figure 4.3: Archival SDSS and follow-up spectroscopy from the subset of the sam-
ple of CL-LINERs new to this work, showing their dramatic changes from LINER
galaxies to AGN.
4.3.3 X-rays
4.3.3.1 Swift
All sources in this sample were observed with the Neil Gehrels Swift Observa-
tory?s (Gehrels et al., 2004) X-ray Telescope (XRT). The XRT data were processed7
(Evans et al., 2009) using HEASOFT v6.228. A subset of the Swift Target of Oppor-
tunity (ToO) observations for AT2018dyk are described in Frederick et al. (2019).
Measured X-ray properties of this sample from multiple epochs taken with XRT on
cadences from months to years are given in Table 4.4.
7http://www.swift.ac.uk/user_objects/
8https://heasarc.gsfc.nasa.gov/docs/software/heasoft/
135
Normalized F? + Constant Normalized F? + Constant
Normalized F? + Constant
4.3.3.2 XMM
We observed a subset of 6 of the sources in this sample with the XMM EPIC
pn camera (Stru?der et al., 2001) during guest observing cycle AO 19 from June to
November in 2020 (program 086505; P.I. Frederick). The XMM-Newton EPIC-pn
observations were taken in Full Frame Window Mode (except for ZTF18aaabltn due
to nearby stellar sources within 20??) and using the Medium Filter for ZTF18aaidlyqand
the Thin Filter for all other sources. The data were reduced using standard tech-
niques with the XMM-Newton Science Analysis System v16.0 (SAS; Gabriel et al.
2004), and EPIC-pn calibration database files updated as of April 2021. We used
a 35?? source extraction region and a 108?? circular source-free background region,
except for ZTF18aaabltn for which we used a 72?? background region, shown in Fig-
ure 4.4. We also adopted CCD event patterns 0 to 4, corresponding to single- and
double-pixel events. The broadband (0.3?10 keV) XMM-Newton light curves of the
six CL-LINERs in this sample are shown in Figure 4.16 the Appendix.
For non-detected sources, we computed upper limits from the XMM EPIC-pn
data at the 3-? level using the XMM SAS/HEASARC command eregionanalyse,
and converted to 0.3-10 keV fluxes with the PIMMS count rate calculator9 assuming
a power law with ?=2, a typical value for AGN. We assessed best-fit models utilizing
?2 statistics and XSPEC version 12.9.1a (Arnaud, 1996). Uncertainties are quoted
at 90% confidence intervals. Model fits and upper limits are given in Table 4.5 in
Section 4.4.1.
Archival XMM-Newton Slew Survey observations of iPTF 16bco indicate the
onset of the X-ray source detected by Swift in its broad-line state to < 1.1 years
before the optical transition observed in (Gezari et al., 2017). ZTF18aasszwr was
previously detected in the XMM Newton Slew Survey with F0.2?12keV = 4.76?10?12
erg cm?2 s?1 (XMM-SSC, 2018). As the only CL-LINER observed in real time
9https://heasarc.gsfc.nasa.gov/cgi-bin/Tools/w3pimms/w3pimms.pl
136
during the ZTF Phase I Survey, AT2018dyk is the only source in the sample in
this follow-up campaign with an archival X-ray spectral observation taken with
XMM-Newton concurrent with its changing-look transition in 2018, 20 days after
the optical peak (PI: Gezari, Program 082204). AT2018dyk was once again faintly
detected by XMM-Newton on 2020 Jun 19, after it had faded in the optical. It
is also the only source in this sample to show significant X-ray variability in the
XMM exposure, exhibiting flaring at the 0.2 cps level during the first 6 ks of the
observation.
4.3.3.3 NuSTAR
NuSTAR observed 4 of the CL-LINERs as part of a joint observing program
with XMM-Newton (program 086505), summarized in Table 4.3. NuSTAR data
were processed (dead-time corrected) with NuSTARDAS version v1.9.2 and with
CALDB updated as of April 2020. We used 50-arcsecond regions to extract the
source events from the NuSTAR images, shown overlaid in Figure 4.5. Images of the
NuSTAR 3?79 keV events in Figure 4.5 show clear high-energy X-ray detections
for 2 of the 4 objects, ZTF18aasszwr and ZTF18aahmkac. Light curves of the
observations are shown in Figure 4.17 of the Appendix. Single focal plane module
(FPM) NuSTAR spectra are shown in Figure 4.7.
137
(a) AT2018dyk (b) ZTF19aambzmf
(c) ZTF18aasszwr (d) ZTF18aahmkac
(e) ZTF18aaabltn (f) ZTF18aaidlyq
Figure 4.4: Source and background extractions regions overlaid on XMM-Newton im-
ages.
138
Table 4.3: NuSTAR Observations of CL LINER candidates
Source z Observation Observation Exposure Count
Name ID Date Time Rate
(UTC) (s) (counts s?1)
ZTF18aajupnt 0.0367 60660001002 2020 Jun 19 21891 <0.03
ZTF19aambzmf 0.122 60660002002 2020 Jul 11 19325 <0.03
ZTF18aasszwr 0.168 60660003002 2020 Dec 01 19762 0.08?0.04
ZTF18aahmkac 0.0699 60660004002 2020 Jun 20 20580 0.09?0.05
(a) AT2018dyk (b) ZTF19aambzmf
(c) ZTF18aasszwr (d) ZTF18aahmkac
Figure 4.5: 50-arcsec source regions overlaid on NuSTAR event file images taken
concurrently with XMM exposures for four sources in the CL-LINER sample. Only
two of the four (ZTF18aasszwr and ZTF18aahmkac) were clearly detected, the other
two (AT2018dyk and ZTF19aambzmf) were not.
139
Table 4.4: Swift XRT Observations of CL LINER Candidates
Source z Observation Observation nH ? Count FX LX
Name ID Date Rate (0.3-10 keV) (0.3-10 keV)
(UTC) (1020) (counts s?1) (10?14) (1042)
ZTF18aaidlyq 0.1000 00011217001 2019 Mar 28 1.72 1.90.2?0.2 0.086? 0.007 1.3? 0.1 3.4? 0.3
ZTF18aasszwr 0.1680 00011216001 2019 Mar 28 1.30 2.00.7?0.6 0.083? 0.011 1.2? 0.2 9.9? 1.3
ZTF18aahiqfi 0.0670 00011214001 2019 Apr 04 1.16 1.60.4?0.3 0.019? 0.003 0.3? 0.1 0.3? 0.1
ZTF18aaabltn 0.0458 00011215001 2019 Mar 25 2.96 2.00.3?0.2 0.176? 0.011 2.9? 0.2 1.5? 0.1
ZTF18aajupnt 0.0367 00095703001 2019 Sep 26 1.59 2.80.3?0.3 0.004? 0.003 0.1? 0.1 0.0? 0.0
ZTF18aahmkac 0.0699 00011466001 2019 Jul 05 1.49 1.80.5?0.3 0.045? 0.005 0.7? 0.1 0.8? 0.1
ZTF18aasuray 0.0397 00011565001 2019 Sep 25 1.11 1.70.6?0.3 0.021? 0.003 0.3? 0.0 0.1? 0.0
ZTF18aaavffc 0.0654 00013203001 2020 Feb 11 1.21 1.80.3?0.2 0.037? 0.004 0.6? 0.1 0.6? 0.1
ZTF19aambzmf 0.1220 00013280001 2020 Mar 05 1.43 1.80.9?0.6 0.012? 0.002 0.4? 0.1 1.6? 0.3
iPTF 16bco 0.2370 00011565001 2016 Oct 21 1.55 2.10.5?0.5 0.027? 0.004 0.4? 0.1 7.0? 1.1
4.4 Analysis
4.4.1 X-rays
For model fitting and calculating luminosities, we used the values for the Galac-
tic column densities and redshifts given in Table 4.5.
140
Table 4.5: XMM EPIC pn Observations of CL LINER candidates
Source z Observation Observation On/Exposurea nb d eH ? Count FX LX
Name ID Date Time Ratec (0.3-10 keV) (0.3-10 keV)
(UTC) (s) (1020) (counts s?1) (10?14) (1042)
AT2018dyk 0.0367 0822040701 2018 Aug 11 1868/11906 1.59 3.02?0.15 0.27?0.01 40?3 1.21?0.09
AT2018dyk 0.0367 0865050101 2020 Jun 19 10492.8/12332 1.59 1.36?0.4 0.023?0.005 6?2 0.18?0.06
ZTF19aambzmf 0.122 0865050301 2020 Jul 09 19423.9/22833 1.43 1.96?0.1 0.075?0.004 15?1 11?1
ZTF18aasszwr 0.168 0865050401 2020 Nov 22 7820.9/9833 1.30 1.8?0.4 0.87?0.01 190?10 148?
ZTF18aahmkac 0.0699 0865050501 2020 Jun 20 17714.5/20668 1.49 1.66?0.02 0.624?0.006 179?3 20?3
ZTF18aaabltn 0.0458 0865050701 2020 Nov 12 7099.7/10284 1.72 1.94?0.03 1.66?0.06 436?7 21.3?0.3
ZTF18aaidlyq 0.1 0865050901 2020 Nov 12 5822.9/6833 1.72 1.87?0.07 0.350?0.008 87?2 21?1
a. Ratio of effective filtered exposure time to total observation duration.
b. Total Galactic column density NH (in units of cm
?2) as a weighted average of measurements from the HI4PI
Collaboration et al. 2016, Kalberla et al. 2005, and Dickey & Lockman 1990.
c. The broadband (03?10 keV) background-subtracted XMM-Newton count rate is given when available. We
report background-subtracted statistical upper limits at the 3-? level.
d. Fluxes (in units of ergs s?1 cm?2) were measured from the best-fit spectral model.
As many of the XMM observations were affected by high levels of background
radiation, only single epochs were taken. Only ZTF19aambzmf shows evidence
of a broad Fe emission feature near 6.4 keV in the spectra shown in Figure 4.6,
approximately 500 days following its discovery in ZTF, although careful modeling
of this potential emission line feature is required. This feature would be expected
in typical broad line AGN, however more time is needed for the narrow Fe emitting
region, located far from the innermost accretion flow and the X-ray continuum
emitting region, to respond to the dramatic continuum change in these sources,
compared to the broad Fe line which shows variations on timescales of less than
1 day (Parker et al., 2016; Vaughan & Edelson, 2001). ZTF18aaabltn is the only
source with a soft excess detected in the XMM follow-up campaign, and is well fit to
a broken power law model with ? = 1.66?0.06 above 2 keV and ? = 2.6?0.2 below
2 keV. All other CL-LINERs were detected and modeled by a simple ? ? 2 power
law, as expected for normal AGN (Brandt et al., 1995), with Galactic absorption
described by the phabs model in XSPEC v12.9.1 (Arnaud, 1996). The majority
showed no indication of a strong Fe line near 6 keV. ZTF18aasszwr is dominated by
background at high energies, and data are therefore only shown below 3 keV.
As expected from the XMM data, power law continua with ? ? 2 are present
141
(a) AT2018dyk (b) ZTF19aambzmf
(c) ZTF18aasszwr (d) ZTF18aahmkac
(e) ZTF18aaabltn (f) ZTF18aaidlyq
Figure 4.6: X-ray spectra of the 6 sources in the sample observed with XMM-Newton
(in chronological order of the exposures): AT2018dyk, ZTF19aambzmf (note the
difference in energy scale), ZTF18aasszwr, ZTF18aahmkac, ZTF18aaabltn, and
ZTF18aaidlyq. The model parameters for Galactic absorption and power law spec-
tral indices are listed in Table 4.5. The top panel of each figure shows the data
modeled by the simple absorbed power law (solid line), except for ZTF18aaabltn
with a soft excess which is modeled by a broken power law (the 2 power law models
are represented as dotted lines). The lower panel of each figure shows the residuals
defined as the ratio of the data over the model, and the background level is indicated
by crosses.
.
142
in the NuSTAR data between 3?79 keV. For the sources in the sample observed
with NuSTAR, emergent features are present at high energies (near 50 keV, close
in energy to where the Compton hump would be expected) at varying strengths in
the X-ray spectra shown in Figure 4.7. We discuss the implications of this further
in Section 4.5. A more detailed analysis of the combined concurrent NuSTAR and
XMM data is forthcoming in a future study.
(a) ZTF18aasszwr (b) ZTF18aahmkac
Figure 4.7: X-ray spectra of the two sources in the sample observed with
NuSTAR concurrently with the XMM-Newton observations, ZTF18aasszwr and
ZTF18aahmkac (note the difference in energy scale). The observation details are
listed in Table 4.3. AT2018dyk and ZTF19aambzmf were not detected above
the background limit, and no clear Compton hump feature is observed in either
ZTF18aasszwr or ZTF18aahmkac above 10 keV as seen in X-ray spectra of normal
AGN.
Following the work of Jin et al. (2021); Ruan et al. (2019a), we compare the
X-ray and UV flux densities of the sources in this sample with other broad line
and changing-look AGN in Figure 4.8. To do so, we measure the simultaneous
optical-to-X-ray flux density ratio (?OX) defined as
?OX = 0.3838 log(L2 keV/L2500A)
by Eq. 4 of Tananbaum et al. (1979), and Eq. 11 of Grupe et al. (2010). Most
sources in this sample are probing the critical inflection point in Eddington ratio
143
which has been shown to be an important condition for triggering the changing-look
transition (see e.g. Noda & Done (2018)). We show the sensitivity of the Eddington
ratio metric to the choice of black hole mass measurement, which affects this result
especially for narrow-line Seyfert 1 sources such as AT2018dyk.
2.2 Low-hard state AGN High-soft state AGN CL-LINERs
(after fading) (before fading) ZTF18aajupnt (MBH,H?)
2.0 ZTF18aajupnt (MBH,?)
1.8 Faint CLQs (Ruan+2019)
Bright CLQs (Ruan+2019)
1.6
BL AGN (Lusso+2010)
1.4
1.2
1.0
0.8
0.6
?4 ?3 ?2 ?1
UV Eddington Ratio log(?L2500A/LEdd)
Figure 4.8: We demonstrate that recent detections are broadly in agreement with,
and are currently probing the critical inversion point in ?OX vs UV Eddington ratio,
the trend established by Ruan et al. (2019a), who scaled state changes of X-ray
binaries (XRBs) to changing-look (also called changing-state) AGN and showed
similar state changes in both. However, we show this trend is sensitive to black
hole mass measurements, which for narrow line sources vary in orders of magnitude
between virial and other methods (see Section 4.4.3). We use this empirical cutoff
to separate the sample into low-hard (blue, left) and high-soft (red, right) states,
analogous to the state changes in XRBs. Shapes used to identify individual sources
are the same as in Figure 4.12.
Figure 4.9 shows the evolution of AT2018dyk along X-ray spectral index and
X-ray Eddington ratio (which is less sensitive to choice of bolometric correction,
see discussion in Ruan et al. 2019a) as measured by all available Swift ToO epochs.
We compare this source to all other CL-LINERs in this sample as well as to 2
X-ray studies of broad-line AGN, Brightman et al. (2013) and Yang et al. 2014,
and find that the sample do not quite approach them in this parameter space,
despite transitioning to broad line quasars in the optical (Frederick et al., 2019).
We speculate on reasons for this discrepancy in Section 4.5.
144
UV/X-ray Spectral Index ?OX
4.0
Yang+2014 BLAGN
Brightman+2013 BLAGN
3.5 CL-LINERs (Swift)
CL-LINERs (XMM)
3.0
2.5
2.0
1.5
1.0
?3.5 ?3.0 ?2.5 ?2.0 ?1.5 ?1.0
X-ray Eddington Ratio log(LX/LEdd)
Figure 4.9: Swift data of AT2018dyk (the source in the sample caught in a real time
transition ?turning-on? and subsequently fading) shows it gradually ?settling? to
where the rest of the sample lies in the parameter space of ?x vs. Lx/LEdd. The
sample in their ?on? states has X-ray power law indices comparable to average broad
line AGN (BLAGN; green and black lines), and undergoes changes in Eddington
ratio between the Swift (2018) and XMM (2020) campaigns.
145
?
We also show the measurements and evolution of the ratio of UV (measured
at 2500 A?) and X-ray (measured at 2 keV) flux densities, ?OX, compared to the
overlapping ranges from populations of LINER galaxies (Maoz, 2007) and broad-line
AGN (Steffen et al., 2006). The majority of the sample lies between the overlapping
region of ?OX within error, and may decrease over time as evidenced by multi-epoch
measurements of AT2018dyk. Although this correlation is weak for the remaining
sources in the sample, this would indicate a relative strengthening of the X-ray
emission while the UV decreases along with the optical, as shown by Frederick et al.
(2019). An increase in the X-ray with time in tidal disruption events is indicative of
the formation of the accretion disk, and may scale with the evolutionary sequence of
XRBs (Wevers et al., 2021), behavior for which there is evidence in some CLAGN
as well (Ruan et al., 2019a).
2.2 LINERs (Maoz+2007) Swift
BLSy1s (Steffen+2006) XMM
2.0
1.8
1.6
1.4
1.2
1.0
0.8
200 400 600 800
Days Since Optical Discovery
Figure 4.10: Ratio of UV to X-ray flux densities, ?OX, of the CL-LINER sample as
measured from Swift UVOT and XRT observations, compared to that of samples of
broad line AGN in red (Steffen et al., 2006) and LINER galaxies in orange (Maoz,
2007).
146
?OX
4.4.2 Optical Spectroscopy
We focus on the broad spectral changes for each source. All follow-up optical
spectra for the sample are summarized in Table 4.7 of the Appendix and selected
spectra are shown in Figure 4.11. Spectra for sources not presented in the Frederick
et al. (2019) sample are shown in Figure 4.3.
Spectra taken with Lowell Discovery Telescope (LDT, formerly DCT) Deveny
spectrograph (PI: Gezari), the Alhambra Faint Object Spectrograph and Camera
(ALFOSC) on the 2.56-m Nordic Optical Telescope (NOT; PI: Sollerman), and
the KAST Double Spectrograph on the Lick 3-m Shane Telescope (PI: Foley) were
reduced using standard IRAF procedures. We performed wavelenth calibration using
arc lamps and flux calibration using a spectrophotometric standard star. Spectra
obtained with the Low-Resolution Imaging Spectrometer (LRIS) on the Keck1 10-
m telescope were reduced automatically with the LRIS reduction pipeline Lpipe
Perley (2019). Spectroscopy obtained with the Double Beam Spectrograph (DBSP)
on the Palomar 200-inch Hale Telescope (P200; PI: Yan) were reduced using the
pyraf-dbsp pipeline (Bellm & Sesar, 2016; Science Software Branch at STScI, 2012).
Data taken by the robotic 2-m Liverpool Telescope (LT) SPectrograph for the Rapid
Acquisition of Transients (SPRAT; PI: Perley) were reduced by the standard pipeline
provided by the Observatorio del Roque de Los Muchachos. Lower resolution (R ?
100) spectra taken by the Palomar 60-inch SEDM were reduced automatically with
pysedm (Rigault et al., 2019).
ZTF18aaidlyq? Both the broad double peaked H? profile and the blue con-
tinuum appear to be fading in the follow-up spectrum taken 3 years post discovery
compared to the spectrum taken 25 days after the discovery in difference imaging.
ZTF18aasuray ? The Fe II complex and broad seemingly double peaked
Balmer emission in this source has remained prominent while the blue continuum
147
[O II] H? H? H? [O III] H? + [N II] [O II] H? H? H? [O III]
8
14 ZTF18aajupnt +393d LDT
7 ZTF18aaidlyq +918d LDT
12 ZTF18aajupnt +335d LDT
6
ZTF18aajupnt +104d LDT
ZTF18aaidlyq +155d LDT 10
5 ZTF18aajupnt +81d Gemini
8
4 ZTF18aajupnt +67d Keck1
ZTF18aaidlyq +25d LDT
6
3 ZTF18aajupnt +51d P60
4 ZTF18aajupnt +11d P200
2
ZTF18aaidlyq -5581d SDSS
2 ZTF18aajupnt -5803d SDSS
1
0
0
3800 4000 4200 4400 4600 4800 5000
4000 4500 5000 5500 6000 6500 7000
Rest Wavelength (A?) Rest Wavelength (A?)
[O II] H? H? H? [O III] H? + [N II]
ZTF18aasuray +940d LDT
5
4 ZTF18aasuray +41d LDT
3
2
ZTF18aasuray -6293d SDSS
1
0
4000 4500 5000 5500 6000 6500 7000
Rest Wavelength (A?)
[O II] H? H? H? [O III] H? + [N II]
12
16bco +422d P200
10
16bco +37d LDT
8
16bco +11d LDT
6
16bco +2d Keck2
4 16bco P60
2 16bco -4368d SDSS
0
4000 4500 5000 5500 6000 6500 7000
Rest Wavelength (A?)
Figure 4.11: Selected multi-epoch spectroscopy from the 3-year follow-up campaign
of the CL-LINER year 1 sample compared to a SDSS spectrum of the host from sev-
eral years prior to the onset of the changing-look transition. Both ZTF18aaidlyq and
ZTF18aasuray (upper left) show potential double peaked emission with a fading in
blue continuum over 3 years, ZTF18aaidlyq shows a fading in H? flux over 3 years,
and AT2018dyk (upper right, with a difference in scale to emphasize the spectrum
in the vicinity of features of interest) shows a fading in H? and He II?4686 over 1
year. Telluric absorption features are indicated by gray shading.
has faded over 2.5 years resulting in the [O III]?5007A?doublet appearing to have
relatively strengthened over that time.
148
Normalized F? + Constant Normalized F? + Constant Normalized F? + Constant
Normalized F? + Constant
He II
AT2018dyk? Spectra taken over 1 year after the optical flare and subsequent
plateau shows a fading in the strong profile of He II?4686A?, and a possible strength-
ening of [O III]?5007A?.
We fit the emission lines of the follow-up spectra with Gaussian models using
lmfit, and present the FWHM measurements in Table 4.6 and Figure 4.14 of the
Appendix. Models of the forbidden lines such as the narrow [N II] emission lines
included in the multi-Gaussian models of the H? profiles are fixed to the width of
the [O III] lines.
We update Figure 14 from Frederick et al. (2019) in Figure 4.12, showing
measurements of the H? and continuum flux ratios of the CL-LINERs compared to
a sample of normal CLQs from MacLeod et al. (2019), to include the latest sources
in this sample as well as additional epochs of the original sample taken months
to years after the initial classification spectra. ZTF18aahmkac is an outlier with
respect to both flux ratios, and ZTF19aambzmf is at the high end of the H? flux
ratios compared to the CLQ sample, although we note the low measurement of the
continuum ratio is due to an issue with the flux calibration in the follow-up spectrum
of ZTF19aambzmf blueward of H?.
We also recreate Figure 17 of Frederick et al. (2019) to include the updated
follow-up data set in Figure 4.13. The emission line model fits to H?, H?, and [O III]
that these measurements are based on are shown in Figure 4.14 in the Appendix,
as well as in the Appendix of Frederick et al. (2019).
149
ZTF18aaabltn
100 ZTF18aahmkac
iPTF 16bco
10 CLQs ZTF18aasuray(MacLeod et al. 2018) ZTF18aahiqfi
AT2018dyk ZTF18aaidlyq
ZTF18aasszwr
1 10 100 1000
fH , high/fH , low 
Figure 4.12: Ratio of continuum flux change as a function of broad line flux change
for our changing-look LINER sample (filled shapes) in comparison to changing-look
AGN (shown in gray). All have much larger (by a factor of > 10) changes in
broad line flux than the changing-look quasar sample. Measurements from a follow-
up epoch of AT2018dyk is shown in orange. The f?3240 ratio measurements are
represented as lower limits, as there is stellar contamination in the low (LINER)
state. We note that there was an error in the flux calibration in the follow-up
spectrum blueward of H? for ZTF19aambzmf, so we do not include the continuum
ratio of that source. Adapted from Figure 6 in MacLeod et al. (2019) and Figure
14 in Frederick et al. (2019).
150
f 3240, high/f 3240, low 
44
18aasszwr iPTF 16bco43
1
42 1
88aaaahhimqfkiac
18aaidlyq 19aambzmf
18aaabltn
41 18aasuray
18aajupnt
4039 40 41 42 43
log ([O III] Luminosity) [erg/s]
Figure 4.13: H? and [O III] ?5007 line luminosities measured for the sample of CL
LINERs, showing that they exist in an extreme region of this parameter space in
the high state. The upper and lower contours representing log LH? vs. log L[O III]
measurements of SDSS DR7 quasars and Sy 1 galaxies from Shen et al. (2011)
and Mullaney et al. (2013) show that this CL-LINER sample is up to an order of
magnitude underluminous in [O III], due to light-travel time delays of an extended
narrow line region that has yet to respond to the continuum flux change. We begin
to see this response of [O III] as H? fades in sources with spectra taken years after
discovery (orange). Fits to these emission lines are shown in Figure 4.15 in the
Appendix. Adapted from Figure 6 in Gezari et al. (2017) and Figure 17 of Frederick
et al. (2019).
4.4.3 Black Hole Mass Estimates
We measure the black hole masses and resulting range of Eddington ratios for
the sample through various methods, presented in Table 4.6, modified from Frederick
et al. (2019) to include the additional sources discovered during the second half of
the ZTF Phase I Survey. The gaussian fits to the H? lines are shown in Figure 4.14.
See the caveats and equations presented in Frederick et al. (2019) for more details.
151
log (H  Luminosity) [erg/s]
Table 4.6: Properties of the host galaxies of the sample of changing-look LINERs
from ZTF and iPTF. MBH for each object is calculated by various methods from the
host galaxy luminosity, mass, and velocity dispersion, respectively, and described in
Section 4.4.3.
g
Name Mar,host log M
b c d e f
Bulge ?F ?L5100A FWHMH log MBH,M log MBH,Bulge log MBH,?F log MBH,vir L/L? r Edd
(mag) [M ] (km s?1 ) (10
43 erg s?1) (km s?1) [M] [M] [M] [M]
ZTF18aahmkac -21.20 10.67 230 0.55?0.01 5599?408 7.6 7.8 7.6 7.4 0.009
ZTF18aaavffc -21.59 10.61 160 2.9?0.2 7020?2060 7.8 7.8 7.7 8.0 0.06
ZTF19aambzmf -21.80 10.74 200 1.13?0.03 1319?318 7.9 7.9 7.9 6.3 0.06
ZTF18aajupnt -22.00 10.66 150 0.49?0.11 939?28 8.0 7.8 7.6 6.4 0.09
ZTF18aasuray -21.70 10.73 230 10.6?0.4 4270?218 7.9 7.9 8.4 8.0 0.03
ZTF18aaidlyq -21.64 - 120 12.7?2.3 7726?458 7.9 - 8.2 8.5 0.06
ZTF18aahiqfi -21.63 - 210 4.1?0.5 8809?723 7.9 - 7.2 8.3 0.2
ZTF18aasszwr -22.19 11.19 180 57.0?1.9 6461?846 8.1 8.3 7.9 8.8 0.5
ZTF18aaabltn -20.62 - 140 0.8?0.2 5195?648 7.3 - 7.5 8.0 0.2
iPTF16bco -22.21 - 176 17.3?11.0 4183?213 8.4 - 7.9 8.1 0.05
a. Computed from the r-band de Vaucouleurs / exponential disk profile model fit magnitude from the SDSS DR14 photometric
catalog, with errors between 0.003?0.005.
b. Computed from broadband SED fits to photometric measurements of SDSS DR 7 galaxies (Mendel et al., 2014), with error on
the measurement of 0.15 dex.
c. Measured from the SDSS spectrum using the PPXF method.
d. McLure & Dunlop (2002)
e. Ha?ring & Rix (2004)
f. Tremaine et al. (2002)
g. In the high state of the source; MBH,?F was employed to obtain the black hole masses used in computing the Eddington ratio).
4.5 Discussion
The detection of hard X-rays in 2 of the 4 sources in this sample observed
with NuSTAR (ZTF18aasszwr and ZTF18aahmkac) indicates the likely existence of
an accretion disk several years after the optical rise associated with the changing-
look transition, as defined in Table 4.2. As there was an X-ray flare observed with
AT2018dyk, this may be evidence for the assembly of some accretion components
occurring concurrently with the changing-look transition observed in 2018 for that
object (Frederick et al., 2019). However, as the host galaxies for the sample are LIN-
ERs, there may have been a pre-existing accretion disk at the time of the changing-
look transition. The physical properties for the persistent LINER accretion flow
may have been importantly different to that of a type 1 AGN, e.g. smaller in phys-
ical extent or lower in density, which may make these a unique physical as well as
observational class of changing-look events, and may explain their more dramatic
nature compared to normal CLAGN (e.g. Figure 4.12).
For the subset of sources with weak or no tentative Compton hump detections
9Measured from the SDSS spectrum using the PPXF method.
152
with NuSTAR, due to the light crossing time from the molecular torus component
being on the order of years, there may not have been sufficient time for any change
following the changing look transition to fully propagate and induce a response.
Fitting the X-ray spectral data with fully physical reflection model is forthcoming
in future work. Additionally, detection of the Compton hump in future X-ray spec-
troscopic observations would be independent support for the (changing-look) AGN
scenario for these nuclear transients, as opposed to other accretion related or flaring
events, as it is a reflection signature of the molecular torus.
In Frederick et al. (2019), the Swift XRT spectrum of AT2018dyk required
an additional strong soft excess component below 1-2 keV modeled by a power law
with ? ? 3, which was no longer seen in the epoch taken 2 years later. Present
in more than 50% of Seyfert 1 galaxies, and prominent in high-mass-accretion-rate
narrow-line Seyfert 1 AGN, it is possible this component is a crucial artifact or
driver of the changing look phenomenon observed in real time in this source. Swift
XRT provides an important insight into the intermediate evolution of the X-ray
spectral slope, shown in Figure 4.9 to approach that of the other objects in the
sample over time. What effects the LINER environment has on the longevity of this
component is an intriguing focus for future theoretical and systematic studies with
a larger sample of CL-LINERs. In tidal disruption events, X-rays are seen to appear
initially from the accretion flow of debris immediately following the destruction of
the star, are sometimes obscured, and later re-emerge in some objects several months
to years later following the formation of the accretion disk (e.g. Gezari et al. 2017;
Wevers et al. 2019).
The fact that AT2018dyk was not strongly detected with XMM-Newton two
years later in 2020 is quite unusual for a NLSy1, which typically display luminous
and variable soft X-rays. It is possible this source is behaving similarly to a non-
canonical X-ray bright TDE that has faded and changed color, despite having a
153
NLSy1-like UV spectrum (the sparse UV spectral data of TDEs are heterogeneous
in the literature; see van Velzen et al. 2020a and references therein). The line
strength of He II, which is directly photoionized by the X-rays and a common but
not ubiquitous spectral signature of TDEs, has also faded with time in this source.
Alternatively, the X-ray emitting region may be partially obscured by Compton-
thick accreted material along the line of sight to the AGN, possibly associated with
the changing-look event.
4.6 Conclusions
As automated discovery and follow-up efforts increase in efficiency, we find
more examples of changing-look AGN which push the limits of what we know of this
population?s properties such as transition timescales and multiwavelength emission
in real time. Such systematic multiwavelength follow up efforts are necessary to
track the evolution of these events in real time. With observations with XMM and
NuSTAR, coupled with multi epoch optical spectroscopy, we attempt to map out
the environments of a growing sample of changing-look LINERs. We find that this
observationally distinct class of objects generally follows trends of X-ray spectral
hardening with time established by Ruan et al. (2019b) and Jin et al. (2021) for
changing-look quasars. We also find a strengthening of the narrow line region tracer
[O III] over several years with respect to the broad line region as expected from a
light travel time delayed response to the dramatic continuum changes observed from
these sources. We observe differences in the directionality of the time evolution of
these sources through various parameter spaces representing accretion rate, hard-
ness, and strengths of emission features tracing various regions of the accretion flow.
Inclusion of intermediate epochs, including from a follow-up campaign with Gem-
ini GMOS-N undertaken between 2019 and 2020, will track this directionality with
finer granularity. Additionally, with statistical samples which will be possible in the
154
near future, these diagnostics may be a way to distinguish heterogeneous accretion
modes or chaotic or unstable phases within this observational class through their
spectroscopic evolution. Systematic searches and follow-up with multi-epoch and
all-sky spectroscopic and X-ray surveys such as SDSS-V and eROSITA will have
the capability to extend this sample to better understand the time evolution of such
events.
4.7 Appendix
Details of the optical and UV spectroscopic follow-up of the CL-LINER sample
are presented in Table 4.7. In Figures 4.15 and 4.14 we present additional Gaussian
profile model fits for selected emission lines of the sample used in Figures 2.17 and
4.12 with methods described in Section 4.4. we present the XMM-Newton and
NuSTAR light curves of the X-ray observed subset of the sample in Figures 4.16
and 4.17, respectively.
155
Name Obs UT Instrument Exposure (s) Reference
ZTF18aajupnt 2002 July 11 SDSS 28816 Abolfathi et al. 2018
2018 June 12 Palomar 200? DBSP 2400 Frederick et al. 2019
2018 July 22 Palomar 60? SED Machine 2430 Frederick et al. 2019
2018 Jul-Dec Swift XRT 40400 Frederick et al. 2019
2018 Aug 7 Keck DEIMOS 300 Frederick et al. 2019
2018 Aug 11 XMM EPIC pn 11906 Frederick et al. 2019
2018 Aug 12 Palomar 60? SED Machine 2430 Frederick et al. 2019
2018 Aug 12 FTN FLOYDS-N 3600 Arcavi et al. 2018
2018 Aug 21 Gemini GMOS-N 600 Frederick et al. 2019
2018 Sept 1 HST STIS 2859 Frederick et al. 2019
2018 Sept 12 DCT Deveny 2400 Frederick et al. 2019
2018 Sept 17 NICER 1856 This work
2019 Jan 18 HST STIS 2502 Frederick et al. 2019
2019 Mar 3 HST STIS 2502 Frederick et al. 2019
2019 Mar 17 Swift 2900 This work
2019 May 2 LDT Deveny 600 This work
2019 Jun 29 LDT Deveny 2400 This work
2019 Sep 26 Swift 1800 This work
2019 Dec 15 Swift 2800 This work
2020 Jan 3 Gemini GMOS-N 1200 This work
ZTF18aasuray 2001 Feb 15 SDSS 8699 Abolfathi et al. 2018
2018 Jun 21 LDT Deveny 1400 Frederick et al. 2019
2019 Sep 25 Swift 2300 This work
2019 Dec 18 Gemini GMOS-N 1200 This work
2020 Dec 6 LDT Deveny 900 This work
ZTF18aahiqfi 2003 Apr 7 SDSS 20314 Abolfathi et al. 2018
2018 Apr 11 LDT Deveny 1800 Frederick et al. 2019
2019 Apr 4 Swift 2000 This work
2019 May 25 Swift 5000 This work
2019 Dec 19 Gemini GMOS-N 1200 This work
ZTF18aaidlyq 2002 Dec 29 SDSS 17642 Abolfathi et al. 2018
2018 May 6 LDT Deveny 1200 Frederick et al. 2019
2019 Mar 28 Swift 2000 This work
2019 Nov 29 Gemini GMOS-N 1200 This work
2020 Oct 15 LDT Deveny 1400 This work
ZTF18aaabltn 2007 Feb 18 SDSS 42982 Abolfathi et al. 2018
2018 Dec 9 Palomar 60? SED Machine 1200 Frederick et al. 2019
2019 Mar 25 Swift 1700 This work
2019 May 2 LDT Deveny 600 Frederick et al. 2019
2019 Nov 29 Gemini GMOS-N 1200 This work
2020 Oct 14 LDT Deveny 1000 This work
ZTF18aasszwr 2003 Jan 5 SDSS 17891 Abolfathi et al. 2018
2018 Dec 3 Palomar 60? SED Machine 2250 Frederick et al. 2019
2019 Mar 28 Swift 857 This work
2019 Nov 5 LDT Deveny 1200 This work
2019 Dec 4 Gemini GMOS-N 1200 This work
2020 Jan 3 Gemini GMOS-N 1200 This work
2021 Mar 3 LDT Deveny 1200 This work
ZTF18aahmkac 2003 Apr 4 SDSS 20200 Abolfathi et al. 2018
2019 May 29 LDT Deveny 900 This work
2019 July 5 Swift 6000 This work
2019 Dec 4 Gemini GMOS-N 1200 This work
ZTF18aaavffc 2005 Apr 12 SDSS 44907 Abolfathi et al. 2018
2020 Jan 29 Gemini GMOS-N 1200 This work
ZTF19aambzmf 2007 May 14 SDSS 44907 Abolfathi et al. 2018
2019 Dec 5 Swift - This work
2020 Jan 3 Gemini GMOS-N 1200 This work
2020 Jan 24 Gemini GMOS-N 1200 This work
iPTF 16bco 2004 June 16 SDSS 27205 Abolfathi et al. 2018
2016 June 2 Palomar 60? SED Machine 1200 Gezari et al. 2017
2016 June 4 Keck DEIMOS 240 Gezari et al. 2017
2016 June 13 LDT Deveny 600 Gezari et al. 2017
2016 July 9 LDT Deveny 1200 Gezari et al. 2017
2016 July 24 Palomar 60? SED Machine 2700 Gezari et al. 2017
2017 July 29 Palomar 200? DBSP 600 This work
2019 Sep 29 Swift 2000 This work
2020 Jan 9 Gemini GMOS-N 1200 This work
2020 Jan 24 Gemini GMOS-N 1200 This work
156
Velocity (km/s) Velocity (km/s) Velocity (km/s)
1e 165000 0 5000 1e 165000 0 50002 1e
401060 2000 0 2000 4000
H ZTF18aahmkac H ZTF18aahmkac [OIII] ZTF18aahmkac
5.0 1.0
2.5 1 0.5
0.0 0 0.0
2.5 FWHM: 5951 ? 135 km s 1 FWHM: 5599 ? 408 km s 1 0.5 FWHM: 797 ? 132 km s
1
6400 6500 6600 6700 4800 4900 5000 4950 5000 5050
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Velocity (km/s) Velocity (km/s) Velocity (km/s)
1e 165000 0 5000 1e401070 2000 0 2000 4000 1e401060 2000 0 2000 4000
H ZTF19aambzmf 4 H ZTF19aambzmf [OIII] ZTF19aambzmf
2
2 2 1
0 0
0
FWHM: 3556 ? 158 km s 1 FWHM: 1319 ? 318 km s 1 1 FWHM: 559 ? 49 km s 1
6400 6500 6600 6700 4800 4850 4900 4950 5000 5050
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Velocity (km/s) Velocity (km/s) Velocity (km/s)
5000 0 5000 1e 165000 0 5000 4000 2000 0 2000 4000
H ZTF18aaavffc H ZTF18aaavffc 100 [OIII] ZTF18aaavffc400
5.0
50
200 2.5
0 0.0
0
FWHM: 4006 ? 149 km s 1 2.5 FWHM: 7021 ? 2062 km s
1 1
50 FWHM: 610 ? 90 km s
6400 6500 6600 6700 4750 4800 4850 4900 4950 5000 4950 5000 5050
Rest Wavelength (?) Rest Wavelength (?) Rest Wavelength (?)
Figure 4.14: Gaussian fits to the H?+[N II], H?, and [O III] line profiles of the latest
sources added to the sample. Measurements from these fits are used in Table 4.6
and Figure 4.13. Note the differences in scale.
157
F  (erg/cm2/s/?)
F  (erg/cm2/s/?) F  (erg/cm2/s/?)
F  (erg/cm2/s/?)
F  (erg/cm2/s/?)
F  (erg/cm2/s/?)
F  (erg/cm2/s/?) F  (erg/cm2/s/?) F  (erg/cm2/s/?)
Velocity (km/s) Velocity (km/s)
1e401060 2000 0 2000 4000 1e401070 2000 0 2000 4000
H ZTF18aajupnt 4 H ZTF18aajupnt
2 2
0 0
FWHM: 2437 ? 159 km s 1 FWHM: 851 ? 278 km s 1
6500 6550 6600 6650 2 4800 4850 4900
Rest Wavelength (?) Rest Wavelength (?)
Velocity (km/s)
1e401060 2000 0 2000 4000
1.5 [OIII] ZTF18aajupnt
1.0
0.5
0.0
0.5 FWHM: 522 ? 88 km s 1
4950 5000 5050
Rest Wavelength (?)
Velocity (km/s) Velocity (km/s)
1e 165000 0 5000 1e 175000 0 5000
H ZTF18aaidlyq H ZTF18aaidlyq
2 4
1 2
0 0
FWHM: 5509 ? 142 km s 1 FWHM: 8807 ? 1152 km s 1
1 2
6400 6500 6600 6700 4750 4800 4850 4900 4950 5000
Rest Wavelength (?) Rest Wavelength (?)
Velocity (km/s)
1e401070 2000 0 2000 4000
6 [OIII] ZTF18aaidlyq
4
2
0
2 FWHM: 440 ? 94 km s 1
4950 5000 5050
Rest Wavelength (?)
Figure 4.15: Gaussian fits to the H?+[N II], H?, and [O III] line profiles using
the latest follow-up data of the original sample displaying spectral changes in Fig-
ure 4.11. Measurements from these fits are used in Table 4.6 and Figure 4.13. Note
the differences in scale.
158
F  (erg/cm2/s/?) F  (erg/cm2/s/?)
F  (erg/cm2/s/?) F  (erg/cm2/s/?)
F  (erg/cm2/s/?) F  (erg/cm2/s/?)
(a) AT2018dyk (b) ZTF19aambzmf
(c) ZTF18aasszwr (d) ZTF18aahmkac
(e) ZTF18aaabltn (f) ZTF18aaidlyq
Figure 4.16: Broadband (0.3?10 keV) light curves of the 6 sources in the sample
observed with XMM-Newton.
.
159
(a) AT2018dyk (b) ZTF19aambzmf
(c) ZTF18aasszwr (d) ZTF18aahmkac
Figure 4.17: Broadband (3?79 keV) finely binned light curves of the 4 sources in
the sample observed with NuSTAR.
.
160
Chapter 5: Conclusions and Future Work
5.1 Lessons Learned
Discoveries of new observational classes of changing-look AGN in the past
decade have uncovered a diverse set of exotic new laboratories for understanding
the relationship between SMBHs at the centers of galaxies and their environments,
specifically how feeding episodes leading to AGN ignition and black hole growth
may proceed differently than we currently understand. Both are processes which
affect how galaxies evolve over cosmological timescales, despite the small relative
size of the central black hole compared to the rest of its host galaxy and the short
timescales of these observations.
In terms of experimental design, ZTF Phase I provided a proof-of-concept
that such new and exciting discoveries could be made in the realm of AGN science
with a survey intended to discover transients such as supernovae. These ambitious
efforts to tease out these events from the alert stream were inevitably met with
difficulties, namely we find that in certain transient events the observational dis-
tinctions between tidal disruption flares, superluminous supernovae, and extreme
AGN variability are subtle, and this should certainly inform how future searches
proceed.
We nonetheless found that these relatively small scale studies of heterogeneous
observational classes of events, of which we conducted intensive follow-up campaigns
and provided detailed accounts across wavebands, were useful for identifying impor-
161
tant characteristics and generating hypotheses for more systematic searches and
studies to eventually test models and theories for exotic accretion-powered phenom-
ena. With this open attention to exciting yet incomplete samples emerging from
various approaches to collecting events from the ZTF alert stream, we found unex-
pected avenues for future study, such as characteristics of nuclei in certain galaxy
subtypes which were far more conducive to hosting such dramatic changes. For
example, Frederick et al. (2019) found a proximity in the BPT diagram to the
cross-over point between LINER and Seyfert galaxies for the sample of CL-LINERs
presented in Chapter 2. Such studies contextualizing these samples within new
theoretical and empirical frameworks are already beginning to be carried out (e.g.
Dodd et al. 2021; Ruan et al. 2019b).
5.2 Future Work
ZTF, along with its wide field survey contemporaries and building off of the
legacy of the Palomar Transient Factory, has revolutionized the field of transient
science beyond even what was intended. Though enabling this work, these pre-
decessors to the Rubin Observatory have opened up a new time domain field of
following up rapid and unique AGN behavior as it occurs, to better and earlier un-
derstand the nature of processes associated with rapid shifts in SMBH accretion on
human timescales, to investigate why they occur.
There exist many interesting future avenues of research to explore, from model-
ing the chemical abundances of the accretion flow with high resolution spectroscopy,
to systematic matched sample studies of the unique host galaxies of these flares,
which will serve to deepen our understanding of the mechanisms driving these new
and exotic events.
For example, Dodd et al. (2021) recently conducted an observational study
of CLAGN host galaxies to understand the conditions for triggering an instability
162
that could result in a change in accretion state, and confirmed this result that
LINERs are conducive to hosting transitions. They found that changing look AGN
are likely linked to periods of galaxy transformation and quenching of star formation
in galaxies undergoing evolution processes that drive gas toward the high-density
nucleus, terminating in episodic bursty accretion as the reservoir is diminished.
5.2.1 Looking Forward to the Landscape of Real-time All-sky and
High Energy Observations
Thanks to technological advancements across instrumentation to data science,
discoveries from the ever-expanding field of multimessenger astrophysics is rich and
growing richer. In the near future (i.e. in the next several years to decades), panoptic
surveys will uncover many more exciting unforeseen avenues for disocvery as well as
systematic exploration into the physical mechanisms underlying these events.
The observational classification scheme presented in Frederick et al. (2020)
will allow for more rapid distinctions to be made between the physical mechanisms
resulting in transients in AGN, to better tailor follow-up efforts in forthcoming
synoptic surveys. Ground-based facilities with multi-epoch imaging photometry
such as PTF/iPTF, ZTF, PanSTARRS, and ASAS-SN have been the workhorses for
photometric surveys discovering and monitoring flaring events of all kinds including
AGN variability, whereas space-based facilities like Kepler have laid the groundwork
for the Transiting Exoplanet Survey Satellite (TESS) to follow AGN with both high
cadence and regular observations (Smith et al., 2018). The Vera Rubin Survey,
formerly LSST, will drastically increase the rate of discovery of candidates for exotic
SMBH accretion events by orders of magnitude, improving our understanding of
their temporal properties on short timescales, and allowing for more rapid follow-
up.
The Sloan Digital Sky Survey, consisting of multi-epoch imaging surveys such
163
as Stripe-821 as well as spectroscopic follow-up like the TDSS variability selected
AGN sample, has been instrumental in defining the catalogues from which CLAGN
may be discovered (e.g. LaMassa et al. 2015; MacLeod et al. 2016; Ruan et al. 2016;
Runnoe et al. 2016). The Black Hole Mapper Project as part of SDSS-V (Kollmeier
et al., 2017) utilizes the legacy of the BOSS fibre multi-object spectrograph and
will provide 800,000 new spectra of more than 6 million total objects with both
Northern and Southern sky coverage as well as spectroscopic monitoring. It joins
large scale IFU surveys such as MUSE to provide shorter cadences (down to 20
minutes) and longer baselines between epochs (up to 20 years), as well as reaching
unprecedented volumetric depths, with a median galaxy redshift of z = 0.1. Large
scale spectroscopic surveys such as those carried out by SDSS-V and the Dark
Energy Spectroscopic Instrument (DESI) will dramatically expand the parameter
space in which changing-state transitions will be found, and thereby growing the
samples and searches of such exotic sources to be both statistical and systematic.
Such samples will allow for tests of the hypotheses put forward here, that certain
environments such as narrow-line Seyfert 1 galaxies and LINER galaxies are more
conducive to hosting these accretion state changes, the most dramatic of this class
seen to-date.
Based on studies of individual objects, it is possible that the accretion mode
changes of CLAGN may be heralded in the X-rays concurrent with (or even prior
to) the optical (e.g. Frederick et al. 2019; Parker et al. 2016; Trakhtenbrot et al.
2019b). The advent of major all-sky X-ray surveys such as eROSITA will comple-
ment rapid follow-up facilities for bright targets such as NICER and Swift XRT,
and will be tracked systematically (completely and homogenously) with large-scale
spectroscopic follow-up campaigns like the SPectroscopic IDentfication of ERosita
Sources (SPIDERS; Clerc et al. 2016) and SDSS-V.
1Stripe-82 was completed between 2000?2008.
164
Complementary to the observational advances made by large scale photometric
and spectroscopic surveys alike, computational advancements in data science as
well as machine learning could be utlized to learn on the photometric datasets
of these now-exotic rapid and smooth AGN flares as they become not-so-unique
when hundreds of thousands are observed throughout the lifetimes of these wide
area surveys, to hone search techniques and identify crucial shared properties of
such events. Future expanded searches may also turn up even more interesting
variants on these flares types, for example SMBH binaries displaying optical flare
counterparts. It will be important for large scale surveys to be positioned to follow
up these binary AGN gravitational wave sources that will be visible to LISA in the
same way as done for stellar mass systems with LIGO. Additionally, there will be
on-going multimessenger searches for neutrinos from TDEs and AGN with IceCube,
which will help us to understand the higher energy emission and processes associated
with SMBH systems accelerating these astroparticles (Stein et al. 2021, in prep.)
5.3 Overview of Contributions
While the number of CLAGN continues to steadily increase, prior to this work
there had yet to be a large-scale systematic study that simultaneously tracks the
appearance of continuum variability and the broad-line emission of changing-look
candidates in real-time using difference imaging as a discovery mechanism. The
Zwicky Transient Facility (ZTF) has allowed for real-time detection and rapid mul-
tiwavelength follow-up of events related to accretion onto supermassive black holes
like never before. Here we presented real-time investigations of extreme transients
associated with sudden changes in the environments of the SMBHs in active galaxy
centers. Possible modes of SMBH accretion ascribed to these outbursts span from
tidal disruption events in active galactic nuclei (AGN), to the curious established
and growing class of ?changing-look? active galactic nuclei (CLAGN). A pilot study
165
with the intermediate Palomar Transient Factory (iPTF) discovered one CL LINER
(Gezari et al., 2017), and ZTF expanded this sample by an order of magnitude,
discovering 9 more (Frederick et al. 2019 and Frederick et al. 2021, in prep.) This
search also resulted in the serendipitous discovery of a sample of 6 new exotic optical
transients in NLSy1s (Frederick et al., 2020), growing the number of such flares by
a factor of three (Trakhtenbrot et al., 2019a).
Filtering the ZTF alert stream for nuclear transients over the past three years
resulted in a surprisingly broad range of accretion powered phenomena associated
with AGN, from flares to changes in accretion state. In studying these extreme
outbursts, we found they were associated with enhanced accretion onto certain types
of supermassive black hole environments. Specifically, we have found a connection
to dramatic flares with spectroscopic changes and LINERs (Chapter 2) and NLSy1s
(Chapter 3), and investigated those new observational classes with multiwavelength
data in real time (Chapter 4).
Through searching for nuclear transients coincident with previously narrow-
line AGN, we discovered eight quiescent galaxies caught ?turning on? into quasars
(Frederick et al., 2019). We showed that the LINER galaxies, classified previously
by weak narrow forbidden line emission in their archival optical spectra, had trans-
formed dramatically into broad-line AGN reminiscent of iPTF 16bco (Gezari et al.,
2017). Comparing the dramatic on-to-off continuum and H? flux ratios with a sam-
ple of CL-Seyferts, we identified that this was a unique class of optical CLAGN
related to physical processes associated with the LINER accretion state. In one
case, follow-up UV and optical spectra revealed the transformation from a LINER
into a narrow-line Seyfert 1 (NLS1) with strong coronal lines. Swift monitoring of
this source revealed bright UV emission that tracked the optical flare, accompanied
by a luminous soft X-ray flare that peaked ?60 days later. Archival light curves
of the sample revealed similar smooth, flare-like deviations from quiescence, and
166
constrained the onset of the optical nuclear flaring (transition timescales shown in
Section 3 Table). This unique class of transients found in ZTF ?turning on? from
quiescent galaxies into broad line AGN on timescales of months to years, promises
to shed light on the nature of LINERs, the nature of coronal line emission, the
nature of the soft X-ray excess in NLSy1s, and the physical mechanism powering
changing-look AGN.
167
Appendix A: Facilities and Software
This thesis work would not have been possible without the following resources:
1. Zwicky Transient Facility
www. ztf. caltech. edu
The ZTF camera is mounted on the Samuel Oschin 48-inch telescope at Palo-
mar Optical Observatory (Bellm et al. 2019a, Graham et al. 2019).
2. The SED Machine
www. ztf. caltech. edu
The SEDM is an IFU spectrograph mounted on the 60-inch telescope at Palo-
mar Optical Observatory.1
3. Lowell Discovery Telescope
https: // lowell. edu/ research/ research-facilities/ 4-3-meter-ldt/
This dissertation relied on data taken by the Deveny Spectrograph at the Lowell
Observatory in Happy Jack, Arizona.
4. SDSS
http: // skyserver. sdss. org/ dr16/ en/ home. aspx
5. SAS
https: // www. cosmos. esa. int/ web/ xmm-newton/ home
1Blagorodnova, N., Neill, J. D., Walters, R., et al. 2018, PASP, 130, 035003. doi:10.1088/1538-
3873/aaa53f
168
Data in this dissertation was reduced using the XMM-Newton Science Analysis
System distributed by ESA.2
6. lmfit
https: // zenodo. org/ record/ 11813# .YDqxgGNOnNY
The lmfit package is Free software, using an MIT license.3,4
7. IPAC, NED
NED. ipac. caltech. edu
This work made use of archival optical spectra made available by the NED
NASA/IPAC Extragalactic Database online interface.5
8. ASASSN Sky Patrol
https: // asas-sn. osu. edu
9. SIMBAD
simbad. u-strasbg. fr/
The SIMBAD database is operated at CDS, Strasbourg, France. 6
2Gabriel et al. 2004
3Newville, Matthew, Stensitzki, Till, Allen, Daniel B., & Ingargiola, Antonino. (2014, Septem-
ber 21). LMFIT: Non-Linear Least-Square Minimization and Curve-Fitting for Python (Version
0.8.0). Zenodo. http://doi.org/10.5281/zenodo.11813
4Newville, M., Stensitzki, T., Allen, D. B., et al. 2016, Astrophysics Source Code Library.
ascl:1606.014
5Madore, Barry, 1998, ?Astrophysics and Algorithms: a DIMACS Workshop on Massive As-
tronomical Data Sets?, meeting abstract, id. 6.
62000,A&AS,143,9 , ?The SIMBAD astronomical database?, Wenger et al.
169
Bibliography
Abolfathi, B., Aguado, D. S., Aguilar, G., et al. 2018, ApJS, 235, 42, doi: 10.3847/
1538-4365/aa9e8a
Agu?eros, M. A., Ivezic?, Z?., Covey, K. R., et al. 2005, AJ, 130, 1022, doi: 10.1086/
432160
Ai, Y. L., Yuan, W., Zhou, H. Y., et al. 2010, ApJL, 716, L31, doi: 10.1088/
2041-8205/716/1/L31
Antonucci, R. 1993, Annual Review of Astronomy and Astrophysics, 31, 473,
doi: 10.1146/annurev.aa.31.090193.002353
Arcavi, I., Trakhtenbrot, B., & Hiramatsu, D. 2019, Transient Name Server As-
troNote, 4, 1
Arcavi, I., Gal-Yam, A., Sullivan, M., et al. 2014, ApJ, 793, 38, doi: 10.1088/
0004-637X/793/1/38
Arcavi, I., Burke, J., French, K. D., et al. 2018, The Astronomer?s Telegram, 11953
Arnaud, K., Branduardi-Raymont, G., Culhane, J., et al. 1985, Monthly Notices of
the Royal Astronomical Society, 217, 105
Arnaud, K. A. 1996, in Astronomical Society of the Pacific Conference Series, Vol.
101, Astronomical Data Analysis Software and Systems V, ed. G. H. Jacoby &
J. Barnes, 17
Assef, R. J., Stern, D., Kochanek, C. S., et al. 2013, ApJ, 772, 26, doi: 10.1088/
0004-637X/772/1/26
Baldi, R. D., Capetti, A., Robinson, A., Laor, A., & Behar, E. 2016, MNRAS, 458,
L69, doi: 10.1093/mnrasl/slw019
Baldwin, J. A., Phillips, M. M., & Terlevich, R. 1981, Publications of the Astro-
nomical Society of the Pacific, 93, 5, doi: 10.1086/130766
170
Barth, A. J., Pancoast, A., Bennert, V. N., et al. 2013, ApJ, 769, 128, doi: 10.1088/
0004-637X/769/2/128
Belfiore, F., Maiolino, R., Maraston, C., et al. 2016, MNRAS, 461, 3111, doi: 10.
1093/mnras/stw1234
Bellm, E. C., & Sesar, B. 2016, pyraf-dbsp: Reduction pipeline for the Palomar
Double Beam Spectrograph. http://ascl.net/1602.002
Bellm, E. C., Kulkarni, S. R., Graham, M. J., et al. 2019a, PASP, 131, 018002,
doi: 10.1088/1538-3873/aaecbe
Bellm, E. C., Kulkarni, S. R., Barlow, T., et al. 2019b, Publications of the Astro-
nomical Society of the Pacific, 131, 068003, doi: 10.1088/1538-3873/ab0c2a
Bianchi, L., Shiao, B., & Thilker, D. 2017, ApJS, 230, 24, doi: 10.3847/1538-4365/
aa7053
Bianchi, L., Seibert, M., Zheng, W., et al. 2005, ApJL, 619, L27, doi: 10.1086/
423710
Blagorodnova, N., Gezari, S., Hung, T., et al. 2017, ApJ, 844, 46, doi: 10.3847/
1538-4357/aa7579
Blagorodnova, N., Neill, J. D., Walters, R., et al. 2018, PASP, 130, 035003, doi: 10.
1088/1538-3873/aaa53f
Blagorodnova, N., Cenko, S. B., Kulkarni, S. R., et al. 2019, ApJ, 873, 92, doi: 10.
3847/1538-4357/ab04b0
Blanchard, P. K., Nicholl, M., Berger, E., et al. 2017, ApJ, 843, 106, doi: 10.3847/
1538-4357/aa77f7
Blandford, R. D., & Znajek, R. L. 1977, MNRAS, 179, 433, doi: 10.1093/mnras/
179.3.433
Boller, T., Brandt, W. N., & Fink, H. 1996, A&A, 305, 53
Bolzonella, M., Miralles, J.-M., & Pello?, R. 2000, A&A, 363, 476
Boroson, T. A., & Green, R. F. 1992, The Astrophysical Journal Supplement Series,
80, 109, doi: 10.1086/191661
Brandt, W. N., Pounds, K. A., & Fink, H. 1995, MNRAS, 273, L47, doi: 10.1093/
mnras/273.1.L47
Bremer, M., Scharwa?chter, J., Eckart, A., et al. 2013, A&A, 558, A34, doi: 10.
1051/0004-6361/201322009
Brightman, M., Silverman, J. D., Mainieri, V., et al. 2013, MNRAS, 433, 2485,
doi: 10.1093/mnras/stt920
171
Brown, J. S., Kochanek, C. S., Holoien, T. W. S., et al. 2018, MNRAS, 473, 1130,
doi: 10.1093/mnras/stx2372
Brown, P. J., Yang, Y., Cooke, J., et al. 2016, ApJ, 828, 3, doi: 10.3847/0004-637X/
828/1/3
Bruch, R. J., Gal-Yam, A., Schulze, S., et al. 2020, arXiv e-prints, arXiv:2008.09986.
https://arxiv.org/abs/2008.09986
Cackett, E. M., Gu?ltekin, K., Bentz, M. C., et al. 2015, ApJ, 810, 86, doi: 10.1088/
0004-637X/810/2/86
Campana, S., Mainetti, D., Colpi, M., et al. 2015, A&A, 581, A17, doi: 10.1051/
0004-6361/201525965
Cappellari, M. 2016, Annual Review of Astronomy and Astrophysics, 54, 597,
doi: 10.1146/annurev-astro-082214-122432
Cappellari, M., & Emsellem, E. 2004, PASP, 116, 138, doi: 10.1086/381875
Cappelluti, N., Predehl, P., Bo?hringer, H., et al. 2011, Memorie della Societa As-
tronomica Italiana Supplementi, 17, 159. https://arxiv.org/abs/1004.5219
Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245, doi: 10.1086/
167900
Cenko, S. B., Fox, D. B., Moon, D.-S., et al. 2006, Publications of the Astronomical
Society of the Pacific, 118, 1396, doi: 10.1086/508366
Cenko, S. B., Cucchiara, A., Roth, N., et al. 2016, ApJL, 818, L32, doi: 10.3847/
2041-8205/818/2/L32
Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, arXiv e-prints,
arXiv:1612.05560. https://arxiv.org/abs/1612.05560
Chan, C.-H., Piran, T., & Krolik, J. H. 2020, arXiv e-prints, arXiv:2004.06234.
https://arxiv.org/abs/2004.06234
Chan, C.-H., Piran, T., Krolik, J. H., & Saban, D. 2019, ApJ, 881, 113, doi: 10.
3847/1538-4357/ab2b40
Chang, Y.-Y., van der Wel, A., da Cunha, E., & Rix, H.-W. 2015, The Astrophysical
Journal Supplement Series, 219, 8, doi: 10.1088/0067-0049/219/1/8
Chatterjee, R., Marscher, A. P., Jorstad, S. G., et al. 2009, The Astrophysical
Journal, 704, 1689
Chen, P. S., Liu, J. Y., & Shan, H. G. 2017, NA, 54, 30, doi: 10.1016/j.newast.
2017.01.005
172
Choi, Y., Gibson, R. R., Becker, A. C., et al. 2014, ApJ, 782, 37, doi: 10.1088/
0004-637X/782/1/37
Cid Fernandes, R., Stasin?ska, G., Mateus, A., & Vale Asari, N. 2011, MNRAS, 413,
1687, doi: 10.1111/j.1365-2966.2011.18244.x
Clerc, N., Merloni, A., Zhang, Y. Y., et al. 2016, MNRAS, 463, 4490, doi: 10.1093/
mnras/stw2214
Constantin, A., & Shields, J. C. 2003, Publications of the Astronomical Society of
the Pacific, 115, 592, doi: 10.1086/374724
Corbett, E. A., Robinson, A., Axon, D. J., et al. 1996, MNRAS, 281, 737, doi: 10.
1093/mnras/281.3.737
Crenshaw, D. M., Kraemer, S. B., & Gabel, J. R. 2003, AJ, 126, 1690, doi: 10.
1086/377625
Crummy, J., Fabian, A., Gallo, L., & Ross, R. 2006, Monthly Notices of the Royal
Astronomical Society, 365, 1067
da Cunha, E., Charlot, S., Dunne, L., Smith, D., & Rowlands, K. 2012, in IAU
Symposium, Vol. 284, The Spectral Energy Distribution of Galaxies - SED 2011,
ed. R. J. Tuffs & C. C. Popescu, 292?296, doi: 10.1017/S1743921312009283
Davis, S. W., Woo, J.-H., & Blaes, O. M. 2007, ApJ, 668, 682, doi: 10.1086/521393
Dawson, K. S., Schlegel, D. J., Ahn, C. P., et al. 2013, AJ, 145, 10, doi: 10.1088/
0004-6256/145/1/10
Decarli, R., Labita, M., Treves, A., & Falomo, R. 2008, MNRAS, 387, 1237, doi: 10.
1111/j.1365-2966.2008.13320.x
Dickey, J. M., & Lockman, F. J. 1990, ARA&A, 28, 215, doi: 10.1146/annurev.
aa.28.090190.001243
Dodd, S. A., Law-Smith, J. A. P., Auchettl, K., Ramirez-Ruiz, E., & Foley, R. J.
2021, ApJL, 907, L21, doi: 10.3847/2041-8213/abd852
Done, C., Davis, S., Jin, C., Blaes, O., & Ward, M. 2012, Monthly Notices of the
Royal Astronomical Society, 420, 1848
Done, C., Davis, S. W., Jin, C., Blaes, O., & Ward, M. 2012, MNRAS, 420, 1848,
doi: 10.1111/j.1365-2966.2011.19779.x
Drake, A. J., Djorgovski, S. G., Mahabal, A., et al. 2009, ApJ, 696, 870, doi: 10.
1088/0004-637X/696/1/870
?. 2011, ApJ, 735, 106, doi: 10.1088/0004-637X/735/2/106
173
Elitzur, M., & Ho, L. C. 2009, ApJ, 701, L91, doi: 10.1088/0004-637X/701/2/L91
Elitzur, M., Ho, L. C., & Trump, J. R. 2014, MNRAS, 438, 3340, doi: 10.1093/
mnras/stt2445
Elvis, M., Wilkes, B. J., McDowell, J. C., et al. 1994, ApJS, 95, 1, doi: 10.1086/
192093
Eracleous, M., & Halpern, J. P. 2001, ApJ, 554, 240, doi: 10.1086/321331
?. 2003, ApJ, 599, 886, doi: 10.1086/379540
Erkens, U., Appenzeller, I., & Wagner, S. 1997, A&A, 323, 707
Evans, P. A., Beardmore, A. P., Page, K. L., et al. 2009, Monthly Notices of the
Royal Astronomical Society, 397, 1177, doi: 10.1111/j.1365-2966.2009.14913.
x
Fabian, A. C., & Ross, R. R. 2010, SSRv, 157, 167, doi: 10.1007/
s11214-010-9699-y
Fazio, G. G., Hora, J. L., Allen, L. E., et al. 2004, The Astrophysical Journal
Supplement Series, 154, 10, doi: 10.1086/422843
Filippenko, A. V. 1996, in Astronomical Society of the Pacific Conference Series,
Vol. 103, The Physics of Liners in View of Recent Observations, ed. M. Eracleous,
A. Koratkar, C. Leitherer, & L. Ho, 17
Forster, K., & Halpern, J. P. 1996, ApJ, 468, 565, doi: 10.1086/177715
Frederick, S., Kara, E., Reynolds, C., Pinto, C., & Fabian, A. 2018, ApJ, 867, 67,
doi: 10.3847/1538-4357/aae306
Frederick, S., Gezari, S., Graham, M. J., et al. 2019, ApJ, 883, 31, doi: 10.3847/
1538-4357/ab3a38
?. 2020, arXiv e-prints, arXiv:2010.08554. https://arxiv.org/abs/2010.08554
Gabriel, C., Denby, M., Fyfe, D., et al. 2004, in Astronomical Data Analysis Software
and Systems (ADASS) XIII, Vol. 314, 759
Galeev, A., Rosner, R., & Vaiana, G. 1979, The Astrophysical Journal, 229, 318
Gallegos-Garcia, M., Law-Smith, J., & Ramirez-Ruiz, E. 2018, ApJ, 857, 109,
doi: 10.3847/1538-4357/aab5b8
Gallo, L. C. 2006, MNRAS, 368, 479, doi: 10.1111/j.1365-2966.2006.10137.x
Garc??a, J. A., Kara, E., Walton, D., et al. 2019, ApJ, 871, 88, doi: 10.3847/
1538-4357/aaf739
174
Gaskell, C. M., & Sparke, L. S. 1986, ApJ, 305, 175, doi: 10.1086/164238
Gaspari, M., & Sadowski, A. 2017, ApJ, 837, 149, doi: 10.3847/1538-4357/aa61a3
Gehrels, N., Chincarini, G., Giommi, P., et al. 2004, ApJ, 611, 1005, doi: 10.1086/
422091
Gelbord, J. M., Mullaney, J. R., & Ward, M. J. 2009, MNRAS, 397, 172, doi: 10.
1111/j.1365-2966.2009.14961.x
George, I. M., & Fabian, A. C. 1991, MNRAS, 249, 352, doi: 10.1093/mnras/249.
2.352
Gezari, S., van Velzen, S., Perley, D., et al. 2019, The Astronomer?s Telegram,
13127, 1
Gezari, S., Chornock, R., Rest, A., et al. 2012, Nature, 485, 217, doi: 10.1038/
nature10990
Gezari, S., Hung, T., Cenko, S. B., et al. 2017, ApJ, 835, 144, doi: 10.3847/
1538-4357/835/2/144
Ghisellini, G. 2013, Radiative processes in high energy astrophysics, Vol. 873
(Springer)
Gierlin?ski, M., & Done, C. 2004, Monthly Notices of the Royal Astronomical Society,
349, L7
Gierlin?ski, M., & Done, C. 2004, MNRAS, 349, L7, doi: 10.1111/j.1365-2966.
2004.07687.x
Goad, M. R., O?Brien, P. T., & Gondhalekar, P. M. 1993, MNRAS, 263, 149, doi: 10.
1093/mnras/263.1.149
Gold. 2019, Galaxies, 7, 63, doi: 10.3390/galaxies7020063
Gonza?lez-Mart??n, O., Masegosa, J., Ma?rquez, I., et al. 2015, A&A, 578, A74, doi: 10.
1051/0004-6361/201425254
Goodrich, R. W. 1989, The Astrophysical Journal, 342, 224
Graham, M. J., Djorgovski, S. G., Drake, A. J., et al. 2017, MNRAS, 470, 4112,
doi: 10.1093/mnras/stx1456
Graham, M. J., Djorgovski, S. G., Stern, D., et al. 2015, MNRAS, 453, 1562, doi: 10.
1093/mnras/stv1726
Graham, M. J., Kulkarni, S. R., Bellm, E. C., et al. 2019, Publications of the As-
tronomical Society of the Pacific, 131, 078001, doi: 10.1088/1538-3873/ab006c
175
Graham, M. J., Ross, N. P., Stern, D., et al. 2020, MNRAS, 491, 4925, doi: 10.
1093/mnras/stz3244
Gromadzki, M., Hamanowicz, A., Wyrzykowski, L., et al. 2019, A&A, 622, L2,
doi: 10.1051/0004-6361/201833682
Grupe, D., Beuermann, K., Mannheim, K., et al. 1995, A&A, 299, L5
Grupe, D., Komossa, S., Leighly, K. M., & Page, K. L. 2010, ApJS, 187, 64, doi: 10.
1088/0067-0049/187/1/64
Grupe, D., Komossa, S., & Saxton, R. 2015, ApJL, 803, L28, doi: 10.1088/
2041-8205/803/2/L28
Grupe, D., Leighly, K. M., Thomas, H.-C., & Laurent-Muehleisen, S. A. 2000, A&A,
356, 11
Grupe, D., & Mathur, S. 2004, ApJL, 606, L41, doi: 10.1086/420975
Guainazzi, M., Nicastro, F., Fiore, F., et al. 1998, MNRAS, 301, L1, doi: 10.1046/
j.1365-8711.1998.02089.x
Guo, H., Sun, M., Liu, X., et al. 2019, ApJL, 883, L44, doi: 10.3847/2041-8213/
ab4138
Halpern, J. P., & Eracleous, M. 1994, ApJL, 433, L17, doi: 10.1086/187537
Ha?ring, N., & Rix, H.-W. 2004, ApJL, 604, L89, doi: 10.1086/383567
Hasinger, G. 2008, A&A, 490, 905, doi: 10.1051/0004-6361:200809839
Hawkins, M. R. S. 1998, A&A, 340, L23. https://arxiv.org/abs/astro-ph/
9810337
?. 2004, Baltic Astronomy, 13, 642
Heckman, T. M. 1980, A&A, 500, 187
Herna?ndez-Toledo, H. M., Va?zquez-Mata, J. A., Mart??nez-Va?zquez, L. A., Choi,
Y.-Y., & Park, C. 2010, AJ, 139, 2525, doi: 10.1088/0004-6256/139/6/2525
HI4PI Collaboration, Ben Bekhti, N., Flo?er, L., et al. 2016, A&A, 594, A116, doi: 10.
1051/0004-6361/201629178
Hills, J. G. 1975, Nature, 254, 295, doi: 10.1038/254295a0
Ho, L. C. 2009, The Astrophysical Journal, 699, 626
Ho, L. C., Filippenko, A. V., & Sargent, W. L. W. 1993, ApJ, 417, 63, doi: 10.
1086/173291
176
?. 1997a, The Astrophysical Journal Supplement Series, 112, 315, doi: 10.1086/
313041
Ho, L. C., Filippenko, A. V., Sargent, W. L. W., & Peng, C. Y. 1997b, The Astro-
physical Journal Supplement Series, 112, 391, doi: 10.1086/313042
Holoien, T. W. S., Brown, J. S., Auchettl, K., et al. 2018, MNRAS, 480, 5689,
doi: 10.1093/mnras/sty2273
Holoien, T. W. S., Kochanek, C. S., Prieto, J. L., et al. 2016a, MNRAS, 455, 2918,
doi: 10.1093/mnras/stv2486
?. 2016b, MNRAS, 463, 3813, doi: 10.1093/mnras/stw2272
Hopkins, P. F., Somerville, R. S., Hernquist, L., et al. 2006, ApJ, 652, 864, doi: 10.
1086/508503
Hung, T., Gezari, S., Blagorodnova, N., et al. 2017, ApJ, 842, 29, doi: 10.3847/
1538-4357/aa7337
Hung, T., Gezari, S., Cenko, S. B., et al. 2018, The Astrophysical Journal Supple-
ment Series, 238, 15, doi: 10.3847/1538-4365/aad8b1
Hung, T., Cenko, S. B., Roth, N., et al. 2019, ApJ, 879, 119, doi: 10.3847/
1538-4357/ab24de
Immler, S., & Lewin, W. H. 2003, in Supernovae and Gamma-Ray Bursters
(Springer), 91?111
Ivezic?, Z?., Kahn, S. M., Tyson, J. A., et al. 2019, ApJ, 873, 111, doi: 10.3847/
1538-4357/ab042c
Jayasinghe, T., Stanek, K. Z., Kochanek, C. S., et al. 2019, MNRAS, 485, 961,
doi: 10.1093/mnras/stz444
Jiang, N., Wang, T., Mou, G., et al. 2019, ApJ, 871, 15, doi: 10.3847/1538-4357/
aaf6b2
Jin, X., Ruan, J. J., Haggard, D., et al. 2021, ApJ, 912, 20, doi: 10.3847/
1538-4357/abeb17
Kalberla, P. M., Burton, W., Hartmann, D., et al. 2005, Astronomy & Astrophysics,
440, 775
Kankare, E., Kotak, R., Mattila, S., et al. 2017, Nature Astronomy, 1, 865, doi: 10.
1038/s41550-017-0290-2
Kara, E., Alston, W. N., Fabian, A. C., et al. 2016, Mon. Not. R. Astron. Soc., 462,
511, doi: 10.1093/mnras/stw1695
177
Kara, E., Pasham, D., Gendreau, K., & Arzoumanian, Z. 2019, The Astronomer?s
Telegram, 13132, 1
Kasliwal, M. M., Cannella, C., Bagdasaryan, A., et al. 2019, PASP, 131, 038003,
doi: 10.1088/1538-3873/aafbc2
Kaspi, S., Smith, P. S., Netzer, H., et al. 2000, ApJ, 533, 631, doi: 10.1086/308704
Kataoka, J., Reeves, J. N., Iwasawa, K., et al. 2007, Publications of the Astronomical
Society of Japan, 59, 279
Kauffmann, G., Heckman, T. M., Tremonti, C., et al. 2003, MNRAS, 346, 1055,
doi: 10.1111/j.1365-2966.2003.07154.x
Kelly, B. C., Bechtold, J., & Siemiginowska, A. 2009, in American Astronomical
Society Meeting Abstracts, Vol. 213, American Astronomical Society Meeting
Abstracts #213, 422.06
Kewley, L. J., Heisler, C. A., Dopita, M. A., & Lumsden, S. 2001, The Astrophysical
Journal Supplement Series, 132, 37, doi: 10.1086/318944
Khachikian, E. Y., & Weedman, D. W. 1971, ApJL, 164, L109, doi: 10.1086/180701
Khazov, D., Yaron, O., Gal-Yam, A., et al. 2016, ApJ, 818, 3, doi: 10.3847/
0004-637X/818/1/3
Kiewe, M., Gal-Yam, A., Arcavi, I., et al. 2012, ApJ, 744, 10, doi: 10.1088/
0004-637X/744/1/10
Klimek, E. S., Gaskell, C. M., & Hedrick, C. H. 2004, ApJ, 609, 69, doi: 10.1086/
420809
Kochanek, C. S. 2016, MNRAS, 458, 127, doi: 10.1093/mnras/stw267
Kochanek, C. S., Shappee, B. J., Stanek, K. Z., et al. 2017, PASP, 129, 104502,
doi: 10.1088/1538-3873/aa80d9
Kollmeier, J. A., Zasowski, G., Rix, H.-W., et al. 2017, arXiv e-prints,
arXiv:1711.03234. https://arxiv.org/abs/1711.03234
Komossa, S. 2002, Reviews in Modern Astronomy, 15, 27. https://arxiv.org/
abs/astro-ph/0209007
Komossa, S., & Bade, N. 1999, A&A, 343, 775. https://arxiv.org/abs/
astro-ph/9901141
Komossa, S., Zhou, H., Wang, T., et al. 2008, ApJ, 678, L13, doi: 10.1086/588281
Komossa, S., Zhou, H., Rau, A., et al. 2009, ApJ, 701, 105, doi: 10.1088/
0004-637X/701/1/105
178
Korista, K., & Ferland, G. 1998, ApJ, 495, 672, doi: 10.1086/305330
Kuminski, E., & Shamir, L. 2016, The Astrophysical Journal Supplement Series,
223, 20, doi: 10.3847/0067-0049/223/2/20
LaMassa, S. M., Urry, C. M., Glikman, E., et al. 2013a, MNRAS, 432, 1351, doi: 10.
1093/mnras/stt553
LaMassa, S. M., Urry, C. M., Cappelluti, N., et al. 2013b, MNRAS, 436, 3581,
doi: 10.1093/mnras/stt1837
LaMassa, S. M., Cales, S., Moran, E. C., et al. 2015, ApJ, 800, 144, doi: 10.1088/
0004-637X/800/2/144
Law, N. M., Kulkarni, S. R., Dekany, R. G., et al. 2009, Publications of the Astro-
nomical Society of the Pacific, 121, 1395, doi: 10.1086/648598
Lawrence, A. 2018, Nature Astronomy, 2, 102, doi: 10.1038/s41550-017-0372-1
Lawrence, A., Gezari, S., Elvis, M., et al. 2012, EPJ Web of Conferences, 39, doi: 10.
1051/epjconf/20123903002
Liu, H.-Y., Yuan, W., Dong, X.-B., Zhou, H., & Liu, W.-J. 2018a, ApJS, 235, 40,
doi: 10.3847/1538-4365/aab88e
Liu, X., Dittmann, A., Shen, Y., & Jiang, L. 2018b, ApJ, 859, 8, doi: 10.3847/
1538-4357/aabb04
Liu, Z., Li, D., Liu, H.-Y., et al. 2020, ApJ, 894, 93, doi: 10.3847/1538-4357/
ab880f
Lohfink, A. M., Reynolds, C. S., Miller, J. M., et al. 2012, The Astrophysical Journal,
758, 67
Lunnan, R., Yan, L., Perley, D. A., et al. 2019, arXiv e-prints, arXiv:1910.02968.
https://arxiv.org/abs/1910.02968
MacLeod, C. L., Ross, N. P., Lawrence, A., et al. 2016, MNRAS, 457, 389, doi: 10.
1093/mnras/stv2997
MacLeod, C. L., Green, P. J., Anderson, S. F., et al. 2019, ApJ, 874, 8, doi: 10.
3847/1538-4357/ab05e2
Mahabal, A., Rebbapragada, U., Walters, R., et al. 2019, Publications of the Astro-
nomical Society of the Pacific, 131, 038002, doi: 10.1088/1538-3873/aaf3fa
Malkan, M. 1983, The Astrophysical Journal, 268, 582
Malyali, A., Rau, A., Arcodia, R., et al. 2020, The Astronomer?s Telegram, 13712,
1
179
Malyali, A., Rau, A., Merloni, A., et al. 2021, A&A, 647, A9, doi: 10.1051/
0004-6361/202039681
Maoz, D. 2007, MNRAS, 377, 1696, doi: 10.1111/j.1365-2966.2007.11735.x
Marconi, A., Axon, D. J., Maiolino, R., et al. 2008, ApJ, 678, 693, doi: 10.1086/
529360
Martin, D. C., Fanson, J., Schiminovich, D., et al. 2005, ApJ, 619, L1, doi: 10.
1086/426387
Marziani, P., & Sulentic, J. W. 2012, New Astronomy Reviews, 56, 49, doi: 10.
1016/j.newar.2011.09.001
Masci, F. J., Laher, R. R., Rebbapragada, U. D., et al. 2017, Publications of the
Astronomical Society of the Pacific, 129, 014002, doi: 10.1088/1538-3873/129/
971/014002
Masci, F. J., Laher, R. R., Rusholme, B., et al. 2019, PASP, 131, 018003, doi: 10.
1088/1538-3873/aae8ac
Mathur, S., Auchettl, K., Kochanek, C. S., et al. 2019, The Astronomer?s Telegram,
13213, 1
Mathur, S., Kuraszkiewicz, J., & Czerny, B. 2001, NA, 6, 321, doi: 10.1016/
S1384-1076(01)00058-6
Mathur, S., Denney, K. D., Gupta, A., et al. 2018, ApJ, 866, 123, doi: 10.3847/
1538-4357/aadd91
McLure, R. J., & Dunlop, J. S. 2002, MNRAS, 331, 795, doi: 10.1046/j.
1365-8711.2002.05236.x
Mendel, J. T., Simard, L., Palmer, M., Ellison, S. L., & Patton, D. R. 2014, ApJS,
210, 3, doi: 10.1088/0067-0049/210/1/3
Merloni, A., & Fabian, A. 2001, Monthly Notices of the Royal Astronomical Society,
321, 549
Merloni, A., Dwelly, T., Salvato, M., et al. 2015, MNRAS, 452, 69, doi: 10.1093/
mnras/stv1095
Meusinger, H., Henze, M., Birkle, K., et al. 2010, A&A, 512, A1, doi: 10.1051/
0004-6361/200913526
Miller, H. R., Ferrara, E. C., McFarland, J. P., et al. 2000, NAR, 44, 539, doi: 10.
1016/S1387-6473(00)00094-4
Miller, J. M., Kaastra, J. S., Miller, M. C., et al. 2015, Nature, 526, 542, doi: 10.
1038/nature15708
180
Miller, J. M., Zoghbi, A., Reynolds, M., et al. 2019, The Astronomer?s Telegram,
13163, 1
Miniutti, G., Ponti, G., Greene, J. E., et al. 2009, MNRAS, 394, 443, doi: 10.1111/
j.1365-2966.2008.14334.x
Molthagen, K., Bade, N., & Wendker, H. J. 1998, A&A, 331, 925
Moriya, T. J., Sorokina, E. I., & Chevalier, R. A. 2018, SSRv, 214, 59, doi: 10.
1007/s11214-018-0493-6
Mullaney, J. R., Alexander, D. M., Fine, S., et al. 2013, MNRAS, 433, 622, doi: 10.
1093/mnras/stt751
Mullaney, J. R., & Ward, M. J. 2008, MNRAS, 385, 53, doi: 10.1111/j.1365-2966.
2007.12777.x
Murayama, T., & Taniguchi, Y. 1998a, ApJ, 503, L115, doi: 10.1086/311554
?. 1998b, ApJ, 497, L9, doi: 10.1086/311264
Nagao, T., Taniguchi, Y., & Murayama, T. 2000, AJ, 119, 2605, doi: 10.1086/
301411
Nandra, K., George, I. M., Mushotzky, R. F., Turner, T. J., & Yaqoob, T. 1997,
ApJ, 476, 70, doi: 10.1086/303600
Nemmen, R. S., Storchi-Bergmann, T., & Eracleous, M. 2014, MNRAS, 438, 2804,
doi: 10.1093/mnras/stt2388
Netzer, H., Mainieri, V., Rosati, P., & Trakhtenbrot, B. 2006, A&A, 453, 525,
doi: 10.1051/0004-6361:20054203
Netzer, H., & Trakhtenbrot, B. 2007, ApJ, 654, 754, doi: 10.1086/509650
Neustadt, J. M. M., Holoien, T. W. S., Kochanek, C. S., et al. 2020, MNRAS, 494,
2538, doi: 10.1093/mnras/staa859
Nikol ajuk, M., Czerny, B., & Gurynowicz, P. 2009, MNRAS, 394, 2141, doi: 10.
1111/j.1365-2966.2009.14478.x
Noda, H., & Done, C. 2018, MNRAS, 480, 3898, doi: 10.1093/mnras/sty2032
Nordin, J., Brinnel, V., van Santen, J., et al. 2019, A&A, 631, A147, doi: 10.1051/
0004-6361/201935634
Nussbaumer, H., & Storey, P. J. 1979, A&A, 74, 244
O?Brien, P. T., Goad, M. R., & Gondhalekar, P. M. 1995, MNRAS, 275, 1125,
doi: 10.1093/mnras/275.4.1125
181
Ofek, E. O., & Frail, D. A. 2011, ApJ, 737, 45, doi: 10.1088/0004-637X/737/1/45
Oknyansky, V. L., Malanchev, K. L., & Gaskell, C. M. 2018, in Revisiting Narrow-
Line Seyfert 1 Galaxies and their Place in the Universe, 12. https://arxiv.org/
abs/1810.08844
Oknyansky, V. L., Gaskell, C. M., Huseynov, N. A., et al. 2016, MNRAS, 000, 1.
https://arxiv.org/abs/arXiv:1701.05042v5
Osterbrock, D. E. 1981, ApJ, 249, 462, doi: 10.1086/159306
Osterbrock, D. E., & Pogge, R. W. 1985, The Astrophysical Journal, 297, 166
Page, K. L., Schartel, N., Turner, M. J., & O?Brien, P. T. 2004, Monthly Notices of
the Royal Astronomical Society, 352, 523
Palaversa, L., Gezari, S., Sesar, B., et al. 2016, ApJ, 819, 151, doi: 10.3847/
0004-637X/819/2/151
Parker, M. L., Komossa, S., Kollatschny, W., et al. 2016, MNRAS, 461, 1927,
doi: 10.1093/mnras/stw1449
Pasham, D., Gendreau, K., Arzoumanian, Z., et al. 2020, The Astronomer?s Tele-
gram, 14036, 1
Patterson, M. T., Bellm, E. C., Rusholme, B., et al. 2019, PASP, 131, 018001,
doi: 10.1088/1538-3873/aae904
Payne, A. V., Shappee, B. J., Hinkle, J. T., et al. 2020, arXiv e-prints,
arXiv:2009.03321. https://arxiv.org/abs/2009.03321
Pelat, D., Alloin, D., & Bica, E. 1987, A&A, 182, 9
Perley, D. A. 2019, PASP, 131, 084503, doi: 10.1088/1538-3873/ab215d
Pfeiffer, M., Appenzeller, I., & Wagner, S. 2000, NAR, 44, 547, doi: 10.1016/
S1387-6473(00)00096-8
Piconcelli, E., Jimenez-Bailo?n, E., Guainazzi, M., et al. 2005, A&A, 432, 15, doi: 10.
1051/0004-6361:20041621
Pogge, R. W. 2000, NAR, 44, 381, doi: 10.1016/S1387-6473(00)00065-8
Poole, T. S., Breeveld, A. A., Page, M. J., et al. 2008, MNRAS, 383, 627, doi: 10.
1111/j.1365-2966.2007.12563.x
Porquet, D., Dumont, A.-M., Collin, S., & Mouchet, M. 1999, A&A, 341, 58
Porquet, D., Reeves, J. N., O?Brien, P., & Brinkmann, W. 2004, A&A, 422, 85,
doi: 10.1051/0004-6361:20047108
182
Pounds, K. A., Done, C., & Osborne, J. P. 1995, MNRAS, 277, L5, doi: 10.1093/
mnras/277.1.L5
Pronik, V. I., & Chuvaev, K. K. 1972, Astrophysics, 8, 112, doi: 10.1007/
BF01002159
Puchnarewicz, E. M., Mason, K. O., Cordova, F. A., et al. 1992, MNRAS, 256, 589,
doi: 10.1093/mnras/256.3.589
Rafter, S. E., Kaspi, S., Chelouche, D., et al. 2013, ApJ, 773, 24, doi: 10.1088/
0004-637X/773/1/24
Rakshit, S., Stalin, C. S., Chand, H., & Zhang, X.-G. 2017, The Astrophysical
Journal Supplement Series, 229, 39, doi: 10.3847/1538-4365/aa6971
Rau, A., Kulkarni, S. R., Law, N. M., et al. 2009, Publications of the Astronomical
Society of the Pacific, 121, 1334, doi: 10.1086/605911
Rigault, M., Neill, J. D., Blagorodnova, N., et al. 2019, A&A, 627, A115, doi: 10.
1051/0004-6361/201935344
Rodr??guez-Ardila, A., Prieto, M. A., Viegas, S., & Gruenwald, R. 2006, ApJ, 653,
1098, doi: 10.1086/508864
Rodr??guez-Ardila, A., Viegas, S. M., Pastoriza, M. G., & Prato, L. 2002, ApJ, 579,
214, doi: 10.1086/342840
Roming, P. W. A., Kennedy, T. E., Mason, K. O., et al. 2005, SSRv, 120, 95,
doi: 10.1007/s11214-005-5095-4
Ross, N. P., Ford, K. E. S., Graham, M., et al. 2018, MNRAS, 480, 4468, doi: 10.
1093/mnras/sty2002
Rowan-Robinson, M. 1977, ApJ, 213, 635, doi: 10.1086/155195
Ruan, J. J., Anderson, S. F., Eracleous, M., et al. 2019a, arXiv e-prints,
arXiv:1903.02553. https://arxiv.org/abs/1903.02553
?. 2019b, arXiv e-prints, arXiv:1909.04676. https://arxiv.org/abs/1909.04676
Ruan, J. J., Anderson, S. F., Cales, S. L., et al. 2016, ApJ, 826, 188, doi: 10.3847/
0004-637X/826/2/188
Rumbaugh, N., Shen, Y., Morganson, E., et al. 2018, ApJ, 854, 160, doi: 10.3847/
1538-4357/aaa9b6
Runnoe, J. C., Cales, S., Ruan, J. J., et al. 2016, MNRAS, 455, 1691, doi: 10.1093/
mnras/stv2385
Ryan, G., & MacFadyen, A. 2017, ApJ, 835, 199, doi: 10.3847/1538-4357/835/2/
199
183
Sarzi, M., Falco?n-Barroso, J., Davies, R. L., et al. 2017, GANDALF: Gas AND
Absorption Line Fitting. http://ascl.net/1708.012
Saxton, C. J., Perets, H. B., & Baskin, A. 2018, MNRAS, 474, 3307, doi: 10.1093/
mnras/stx2928
Schawinski, K., Thomas, D., Sarzi, M., et al. 2007, MNRAS, 382, 1415, doi: 10.
1111/j.1365-2966.2007.12487.x
Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103, doi: 10.1088/0004-637X/
737/2/103
Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525, doi: 10.1086/
305772
Schmidt, R. W., & Wambsganss, J. 2010, General Relativity and Gravitation, 42,
2127, doi: 10.1007/s10714-010-0956-x
Schneider, D. P., Richards, G. T., Hall, P. B., et al. 2010, AJ, 139, 2360, doi: 10.
1088/0004-6256/139/6/2360
Science Software Branch at STScI. 2012, PyRAF: Python alternative for IRAF.
http://ascl.net/1207.011
Shappee, B. J., Prieto, J. L., Grupe, D., et al. 2014, ApJ, 788, 48, doi: 10.1088/
0004-637X/788/1/48
Shen, Y., & Ho, L. C. 2014, Nature, 513, 210, doi: 10.1038/nature13712
Shen, Y., Richards, G. T., Strauss, M. A., et al. 2011, ApJS, 194, 45, doi: 10.1088/
0067-0049/194/2/45
Sheng, Z., Wang, T., Jiang, N., et al. 2017, ApJ, 846, L7, doi: 10.3847/2041-8213/
aa85de
Siemiginowska, A., Czerny, B., & Kostyunin, V. 1996, ApJ, 458, 491, doi: 10.1086/
176831
Silk, J., & Rees, M. J. 1998, A&A, 331, L1. https://arxiv.org/abs/astro-ph/
9801013
Singh, K., Garmire, G., & Nousek, J. 1985, The Astrophysical Journal, 297, 633
Singh, K. P., Garmire, G. P., & Nousek, J. 1985, ApJ, 297, 633, doi: 10.1086/163560
Singh, R., van de Ven, G., Jahnke, K., et al. 2013, A&A, 558, A43, doi: 10.1051/
0004-6361/201322062
Smartt, S. J., Smith, K. W., McBrien, O., et al. 2019, Transient Name Server
AstroNote, 33, 1
184
Smith, K. L., Mushotzky, R. F., Boyd, P. T., et al. 2018, ApJ, 857, 141, doi: 10.
3847/1538-4357/aab88d
Smith, N., Silverman, J. M., Chornock, R., et al. 2009, ApJ, 695, 1334, doi: 10.
1088/0004-637X/695/2/1334
Sollerman, J., Taddia, F., Arcavi, I., et al. 2019, A&A, 621, A30, doi: 10.1051/
0004-6361/201833689
Sollerman, J., Fransson, C., Barbarino, C., et al. 2020, arXiv e-prints,
arXiv:2009.04154. https://arxiv.org/abs/2009.04154
Soumagnac, M. T., & Ofek, E. O. 2018, PASP, 130, 075002, doi: 10.1088/
1538-3873/aac410
Steffen, A. T., Strateva, I., Brandt, W. N., et al. 2006, AJ, 131, 2826, doi: 10.1086/
503627
Stein, R., van Velzen, S., Kowalski, M., et al. 2020, arXiv e-prints, arXiv:2005.05340.
https://arxiv.org/abs/2005.05340
Stephens, S. A. 1989, AJ, 97, 10, doi: 10.1086/114953
Stern, D., Assef, R. J., Benford, D. J., et al. 2012, ApJ, 753, 30, doi: 10.1088/
0004-637X/753/1/30
Stern, D., McKernan, B., Graham, M. J., et al. 2018, ApJ, 864, 27, doi: 10.3847/
1538-4357/aac726
Stern, J., & Laor, A. 2012, MNRAS, 423, 600, doi: 10.1111/j.1365-2966.2012.
20901.x
Storchi-Bergmann, T., Baldwin, J. A., & Wilson, A. S. 1993, ApJL, 410, L11,
doi: 10.1086/186867
Stru?der, L., Briel, U., Dennerl, K., et al. 2001, Astronomy & Astrophysics, 365, L18
Tadhunter, C., Spence, R., Rose, M., Mullaney, J., & Crowther, P. 2017, Nature
Astronomy, 1, 0061, doi: 10.1038/s41550-017-0061
Tananbaum, H., Avni, Y., Branduardi, G., et al. 1979, ApJL, 234, L9, doi: 10.
1086/183100
Terlevich, R., Tenorio-Tagle, G., Franco, J., & Melnick, J. 1992, MNRAS, 255, 713,
doi: 10.1093/mnras/255.4.713
Thomas, D., Steele, O., Maraston, C., et al. 2013, MNRAS, 431, 1383, doi: 10.
1093/mnras/stt261
Tonry, J. L., Denneau, L., Heinze, A. N., et al. 2018, Publications of the Astronom-
ical Society of the Pacific, 130, 064505, doi: 10.1088/1538-3873/aabadf
185
Trakhtenbrot, B., Arcavi, I., Ricci, C., & Burke, J. 2020, Transient Name Server
AstroNote, 105, 1
Trakhtenbrot, B., Arcavi, I., Ricci, C., et al. 2019a, Nature Astronomy, 3, 242,
doi: 10.1038/s41550-018-0661-3
Trakhtenbrot, B., Arcavi, I., MacLeod, C. L., et al. 2019b, arXiv e-prints,
arXiv:1903.11084. https://arxiv.org/abs/1903.11084
Tremaine, S., Gebhardt, K., Bender, R., et al. 2002, ApJ, 574, 740, doi: 10.1086/
341002
Turner, T., & Pounds, K. 1989, Monthly Notices of the Royal Astronomical Society,
240, 833
Urry, C. M., & Padovani, P. 1995, Publications of the Astronomical Society of the
Pacific, 107, 803, doi: 10.1086/133630
Uttley, P., Cackett, E. M., Fabian, A. C., Kara, E., & Wilkins, D. R. 2014, Astron-
omy and Astrophysics Reviews, 22, 72, doi: 10.1007/s00159-014-0072-0
Uttley, P., McHardy, I. M., Papadakis, I. E., Guainazzi, M., & Fruscione, A. 1999,
MNRAS, 307, L6, doi: 10.1046/j.1365-8711.1999.02801.x
van Velzen, S., Holoien, T. W. S., Onori, F., Hung, T., & Arcavi, I. 2020a, arXiv
e-prints, arXiv:2008.05461. https://arxiv.org/abs/2008.05461
van Velzen, S., Mendez, A. J., Krolik, J. H., & Gorjian, V. 2016, ApJ, 829, 19,
doi: 10.3847/0004-637X/829/1/19
van Velzen, S., Farrar, G. R., Gezari, S., et al. 2011, ApJ, 741, 73, doi: 10.1088/
0004-637X/741/2/73
van Velzen, S., Gezari, S., Cenko, S. B., et al. 2019, ApJ, 872, 198, doi: 10.3847/
1538-4357/aafe0c
van Velzen, S., Gezari, S., Hammerstein, E., et al. 2020b, arXiv e-prints,
arXiv:2001.01409. https://arxiv.org/abs/2001.01409
Vaughan, S., & Edelson, R. 2001, ApJ, 548, 694, doi: 10.1086/319028
Ve?ron-Cetty, M. P., & Ve?ron, P. 2000, Astronomy and Astrophysics Reviews, 10,
81, doi: 10.1007/s001590000006
Ve?ron-Cetty, M.-P., Ve?ron, P., & Gonc?alves, A. C. 2001, A&A, 372, 730, doi: 10.
1051/0004-6361:20010489
Voges, W., Aschenbach, B., Boller, T., et al. 1999, A&A, 349, 389. https://arxiv.
org/abs/astro-ph/9909315
186
Wang, T., Brinkmann, W., & Bergeron, J. 1996, A&A, 309, 81
Wang, T.-G., Zhou, H.-Y., Komossa, S., et al. 2012, ApJ, 749, 115, doi: 10.1088/
0004-637X/749/2/115
Wang, T.-G., Zhou, H.-Y., Wang, L.-F., Lu, H.-L., & Xu, D. 2011, ApJ, 740, 85,
doi: 10.1088/0004-637X/740/2/85
Wevers, T., Pasham, D. R., van Velzen, S., et al. 2019, MNRAS, 488, 4816, doi: 10.
1093/mnras/stz1976
?. 2021, arXiv e-prints, arXiv:2101.04692. https://arxiv.org/abs/2101.04692
Wilkins, D. R., Gallo, L. C., Grupe, D., et al. 2015, MNRAS, 454, 4440, doi: 10.
1093/mnras/stv2130
Winkler, H. 1992, MNRAS, 257, 677, doi: 10.1093/mnras/257.4.677
Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868,
doi: 10.1088/0004-6256/140/6/1868
XMM-SSC. 2018, VizieR Online Data Catalog, IX/53
Xu, D., Komossa, S., Zhou, H., et al. 2012, AJ, 143, 83, doi: 10.1088/0004-6256/
143/4/83
Yan, L., Wang, T., Jiang, N., et al. 2019, ApJ, 874, 44, doi: 10.3847/1538-4357/
ab074b
Yan, L., Perley, D., Schulze, S., et al. 2020, arXiv e-prints, arXiv:2006.13758. https:
//arxiv.org/abs/2006.13758
Zhou, H., Wang, T., Yuan, W., et al. 2006, ApJS, 166, 128, doi: 10.1086/504869
187