ABSTRACT
Title of dissertation: THE COOL SIDE OF GALACTIC WINDS:
EXPLORATION WITH HERSCHEL-PACS
AND SPITZER-IRS
Myra Jean Stone, Doctor of Philosophy, 2020
Dissertation directed by: Professor Sylvain Veilleux
Department of Astronomy
Galactic-scale outflows driven by starbursts and/or active galactic nuclei (AGN)
are key ingredients to theoretical models and numerical simulations of galaxy as-
sembly and evolution. The feedback induced by the presence of these outflows (or
winds) may affect the evolution and formation of a galaxy by regulating the amount
of cold, dense gas responsible for star formation and black hole accretion.
We present the results from a systematic search for galactic-scale, molecu-
lar (OH 119 ?m) outflows in a sample of 52 Local Volume (d < 50 Mpc) Burst
Alert Telescope detected active galactic nuclei (BAT AGN) with Herschel -PACS.
We combine the results from our analysis of the BAT AGN with the published
Herschel/PACS data of 43 nearby (z < 0.3) galaxy mergers, mostly ultraluminous
infrared galaxies (ULIRGs) and QSOs. Our data show that both the starburst and
AGN contribute to driving OH outflows, but the fastest OH winds require AGN
with quasar-like luminosities.
We also analyze Spitzer InfraRed Spectrograph (IRS) observations of the
OH 35 ?m feature in 15 nearby (z . 0.06) (ultra-)luminous infrared galaxies
(U/LIRGs). The measured OH 35 ?m equivalent widths are used to compute an
average OH column density which is then compared to the hydrogen column density
for a typical optical depth at 35 ?m of ?0.5 and gas-to-dust ratio of 125 to derive
an OH?to?H abundance ratio of XOH = 1.01 ? 0.15 ? 10?6. The OH 35 ?m line
profiles predicted from published radiative transfer models constrained by observa-
tions of OH 65, 79, 84, and 119 ?m in five objects are found to be consistent with
the IRS OH 35 ?m spectra.
Finally, we analyze Herschel -PACS observations of five atomic fine-structure
transition lines ([O I] 63 ?m, [O III] 88 ?m, [N II] 122 ?m, [O I] 145 ?m, and
[C II] 158 ?m) in seven nearby (d < 16 Mpc) galaxies with well-known galactic-
scale outflows (Cen A, Circinus, M 82, NGC 253, NGC 1068, NGC 3079, and NGC
4945). With this suite of atomic emission lines, we investigate the cool neutral
atomic (T ? 103 K) and warm ionized (T ? 104 K) gas phases within each outflow.
The outflows in the Herschel data are spatially isolated from the galactic disk based
on the kinematic signatures of the outflows. The spatial distribution and physical
properties of the outflows detected in the Herschel data are compared with published
results at other wavelengths. For completeness, an analysis of the molecular gas
traced by OH 119 ?m is also presented.
THE COOL SIDE OF GALACTIC WINDS:
EXPLORATION WITH HERSCHEL-PACS AND SPITZER-IRS
by
Myra Jean Stone
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
2020
Advisory Committee:
Professor Sylvain Veilleux, Chair/Advisor
Professor Alberto Bolatto
Professor Bill Dorland
Dr. Marcio Mele?ndez
Professor Stuart Vogel
?c Copyright by
Myra Jean Stone
2020

Preface
Much of the work in this thesis has been published and is presented here with
minimal modification and the rest is in preparation for submission.
Chapter 2 has been published in The Astrophysical Journal under the title
?The Search for Molecular Outflows in Local Volume AGNs with Herschel-PACS?
(Stone, M., Veilleux, S., Mele?ndez, M., Sturm, E., Gracia?-Carpio, J., & Gonza?lez-
Alfonso, E., 2016, ApJ, 826,111).
Chapter 3 has been published in The Astrophysical Journal under the title
?Constraints on the OH-to-H Abundance Ratio in Infrared-bright Galaxies Derived
from the Strength of the OH 35 ?m Absorption Feature? (Stone, M., Veilleux, S.,
Gonza?lez-Alfonso, E., Spoon, H., & Sturm, E., 2018, ApJ, 853, 132).
Chapter 4 is currently in preparation for submission to The Astrophysical
Journal (Stone, M. & Veilleux, S.).
ii
Acknowledgments
I would not have made it this far and completed this thesis without the im-
mense support from the people around me. I am very grateful to my advisor, Sylvain
Veilleux, for his advice and guidance in my research and his enduring ability to keep
me from getting too far down into many a rabbit hole.
I am indebted to Marcio Mele?ndez for always being available to chat about
life, the universe, and everything (mostly universe related though). His invaluable
assistance with my research and coding throughout my entire graduate career was
one of the main propellers that kept me moving forward. His obvious excitement
about astronomy and his cheerful eagerness to share his knowledge about it has
always been a reminder to me as to why I too loved the field.
I thank Blake and Zeeve for our many take-out dinners and TV watching after
long days of work. These outings always kept me sane ... and looking forward
to food. Their support both personal and professional has meant a great deal to
me. I would like to thank Alex who was a constant source of advice, laughter, and
unwavering support which kept me afloat early on in these uncharted waters. I?d
like to thank Qian and Tiara for our food adventures while seeking out new and
interesting restaurants. I thank Drew for our many chats which helped me navigate
the ins and outs of the department and for the many chats which just helped brighten
my mood.
I would like to thank my old roommates Tanya and Paulo who endured with
me a very interesting 2.5 year stint in an apartment whose description would cause
iii
your brow to furrow and your eyes to squint. I?m grateful for our little traditions
such as RenFest and BBQ, Blues, and Brews which were always great ways to relax
and reset. To my more recent roommates, Aaron, Lauren, Silvy and Chris, I am
grateful for the welcoming home we made which was such a comfort during hard
times. I?m grateful for our dinners and cocktail hours which kept me fueled for
another day of research.
Last but not least, I?d like to thank my boyfriend Chris who has been amazing
during this last leg of my graduate school journey. You?ve pushed me to strive for
the best and set me back on course if ever I faltered. Your belief in me has helped
me maintain belief in myself, something for which I will always be grateful for.
iv
Table of Contents
Preface ii
Acknowledgements iii
Table of Contents v
List of Tables viii
List of Figures ix
List of Abbreviations xvi
1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Starburst and Active Galaxies . . . . . . . . . . . . . . . . . . . . . . 5
1.2.1 Starburst-Driven Winds . . . . . . . . . . . . . . . . . . . . . 5
1.2.2 AGN-Driven Winds . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 The Multiphase ISM in Outflows . . . . . . . . . . . . . . . . . . . . 11
1.3.1 General Considerations . . . . . . . . . . . . . . . . . . . . . . 11
1.3.2 The OH Molecule . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3.3 Atomic Fine-Structure Lines . . . . . . . . . . . . . . . . . . . 15
1.4 Observatories and Instruments . . . . . . . . . . . . . . . . . . . . . . 17
1.4.1 Herschel -PACS . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.4.2 Spitzer -IRS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.5 Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2 Search for Molecular Outflows in Local Volume AGN with Herschel -PACS 22
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2 BAT AGN Sample Selection . . . . . . . . . . . . . . . . . . . . . . . 29
2.3 ULIRG and PG QSO Sample . . . . . . . . . . . . . . . . . . . . . . 36
2.4 Properties of the Sample Galaxies . . . . . . . . . . . . . . . . . . . . 36
2.5 Observations, Data Reduction, and Spectral Analysis . . . . . . . . . 44
2.5.1 OH 119 ?m Feature . . . . . . . . . . . . . . . . . . . . . . . . 44
2.5.1.1 OH Observations . . . . . . . . . . . . . . . . . . . . 44
2.5.1.2 OH Data Reduction . . . . . . . . . . . . . . . . . . 45
2.5.1.3 Spectral Analysis of the OH Doublet . . . . . . . . . 46
2.5.2 The 9.7 ?m Silicate Feature . . . . . . . . . . . . . . . . . . . 49
v
2.5.2.1 Data Reduction of the 9.7 ?m Silicate Feature . . . . 49
2.5.2.2 Spectral Analysis of the 9.7 ?m Silicate Feature . . . 50
2.6 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.6.1 The OH 119 ?m Feature . . . . . . . . . . . . . . . . . . . . . 52
2.6.2 The S9.7?m Feature . . . . . . . . . . . . . . . . . . . . . . . . 58
2.6.3 OH Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.7 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.7.1 Outflows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.7.1.1 BAT AGN with Molecular Outflows . . . . . . . . . 70
2.7.1.2 Driving Mechanisms of Molecular Outflows . . . . . 74
2.7.2 Inflows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.7.3 The 9.7 ?m Silicate Feature . . . . . . . . . . . . . . . . . . . 78
2.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3 Constraints on the OH-to-H Abundance Ratio in Infrared-Bright Galaxies
Derived from the Strength of the OH 35 ?m Absorption Feature 85
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.2 Background: The OH Molecule . . . . . . . . . . . . . . . . . . . . . 90
3.3 Sample Selection, Data Reduction, and Spectral Analysis . . . . . . . 91
3.3.1 The Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.3.2 Data Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.3.3 Spectral Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
3.4.1 OH Column Density . . . . . . . . . . . . . . . . . . . . . . . 102
3.4.2 XOH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
3.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
3.5.1 XOH: Comparison with the Literature . . . . . . . . . . . . . 106
3.5.2 XOH: A Check on Radiative Transfer Models . . . . . . . . . . 107
3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
4 Far-Infrared Integral-Field Spectroscopy of Nearby
Galactic Winds with Herschel-PACS 110
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4.2 Atomic Fine-Structure Emission Lines . . . . . . . . . . . . . . . . . 113
4.3 The Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.4 Observations, Archival Data, and Spectral Analysis . . . . . . . . . . 126
4.5 Velocity-Integrated Emission Line Flux and Ratio Maps . . . . . . . . 128
4.5.1 Constraints from the Emission Line Ratios . . . . . . . . . . . 129
4.5.2 Results from the Emission Line Ratios . . . . . . . . . . . . . 130
4.6 Gas Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
4.7 PDR Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
4.7.1 [C II] 158 Emission . . . . . . . . . . . . . . . . . . . . . . . . 142
4.7.2 [O I] 63 Emission . . . . . . . . . . . . . . . . . . . . . . . . . 142
4.7.3 Total Infrared Emission . . . . . . . . . . . . . . . . . . . . . 143
4.7.4 Hydrogen Density and UV Radiation Field . . . . . . . . . . . 144
vi
4.7.5 Caveats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
4.8 Methods to Derive the Outflow Properties . . . . . . . . . . . . . . . 146
4.8.1 Beam Smearing . . . . . . . . . . . . . . . . . . . . . . . . . . 146
4.8.2 Tilted Ring Model . . . . . . . . . . . . . . . . . . . . . . . . 147
4.8.3 Defining the Spatial Location of the Outflow . . . . . . . . . . 148
4.8.4 Line Luminosities . . . . . . . . . . . . . . . . . . . . . . . . . 150
4.8.5 Mass of the Neutral Atomic Wind . . . . . . . . . . . . . . . . 151
4.8.6 Mass of the Ionized Wind . . . . . . . . . . . . . . . . . . . . 152
4.8.7 Kinetic Energy of the Wind . . . . . . . . . . . . . . . . . . . 153
4.9 Properties of the Outflows and Multi-Phase Comparisons . . . . . . . 153
4.9.1 M 82 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
4.9.2 Cen A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
4.9.3 Circinus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
4.9.4 NGC 253 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
4.9.5 NGC 1068 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4.9.6 NGC 3079 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.9.7 NGC 4945 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
4.10 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
5 Summary and Future Work 187
5.1 Main Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
A Spitzer MIR SPECTRA 192
B Results on OH 119 ?m 199
Bibliography 222
vii
List of Tables
2.1 Herschel Observations of BAT AGN . . . . . . . . . . . . . . . . . . 23
2.1 Herschel Observations of BAT AGN . . . . . . . . . . . . . . . . . . 24
2.2 Galaxy Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2 Galaxy Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2 Galaxy Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.3 Properties of the OH 119 ?m Profiles . . . . . . . . . . . . . . . . . 39
2.3 Properties of the OH 119 ?m Profiles . . . . . . . . . . . . . . . . . 40
2.3 Properties of the OH 119 ?m Profiles . . . . . . . . . . . . . . . . . 41
2.4 S9.7?m Continuum Parameters and Measured Values of the BAT AGN
Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.4 S9.7?m Continuum Parameters and Measured Values of the BAT AGN
Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.4 S9.7?m Continuum Parameters and Measured Values of the BAT AGN
Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.5 S9.7?m Continuum Parameters and Measured Values of the ULIRG/PG
QSO Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.5 S9.7?m Continuum Parameters and Measured Values of the ULIRG/PG
QSO Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.6 Results from Statistical Analyses of Host Galaxy Properties . . . . . 64
2.7 Results from Statistical Analyses of the Kinematics . . . . . . . . . . 65
2.7 Results from Statistical Analyses of the Kinematics . . . . . . . . . . 66
3.1 Spitzer -IRS Spectra: Observations . . . . . . . . . . . . . . . . . . . 88
3.2 Galaxy Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.3 OH 35 ?m Profile Properties . . . . . . . . . . . . . . . . . . . . . . 101
4.1 Fine-structure Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.2 Galaxy Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
4.3 PACS Fine-Structure Line Observations . . . . . . . . . . . . . . . . 121
4.3 PACS Fine-Structure Line Observations . . . . . . . . . . . . . . . . 122
4.4 Wind Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
B.1 PACS OH 119 ?m Observations . . . . . . . . . . . . . . . . . . . . 200
viii
List of Figures
1.1 SMBH vs galaxy properties . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Schematic of two evolutionary phases of a wind-blown bubble . . . . 8
1.3 Schematic of a shock driven into the ISM by an AGN . . . . . . . . . 10
1.4 Multiphase AGN Outflow . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Multiphase SB Outflow in M 82 . . . . . . . . . . . . . . . . . . . . . 14
1.6 Grotrian diagram of OH ground state transitions . . . . . . . . . . . 16
1.7 Grotrian diagram of the electronic transitions in atomic fine-structure
lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.8 Herschel, Spitzer, model SEDs . . . . . . . . . . . . . . . . . . . . . . 18
2.1 BAT/AGN and ULIRG/PG QSO property distributions . . . . . . . 24
2.2 Fits to the continuum subtracted OH 119 line profiles . . . . . . . . . 27
2.2 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.2 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.2 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.3 OH 119 EW vs M? vs ?AGN vs LAGN . . . . . . . . . . . . . . . . . . 45
2.4 The apparent strength of the 9.7 ?m silicate feature relative to the
local mid-infrared continuum as a function of the AGN fractions.
Note that S9.7?m is a logarithmic quantity and can be interpreted as
the apparent silicate optical depth. The strength of silicate absorption
increases upward. Sign conventions and meanings of the symbols are
the same as those in Section 2.5.1.1. Crosses represent objects with a
null OH detection. Vertical lines indicate objects with a null 9.7 ?m
silicate feature detection. . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.5 Title . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
ix
2.6 Histograms showing the distributions of the 50% (median; left panels)
and 84% (right panels) velocities derived from the multi-Gaussian fits
to the OH profiles of the BAT AGN. Top panels show pure absorption
components (blue), pure emission components (red dashed), P-cygni
emission components (red, left diagonals), and inverse P-cygni emis-
sion components (red, right diagonals). Bottom panels show pure
absorption components (filled grey), P-Cygni absorption components
(blue, left diagonals), inverse P-cygni absorption components (blue,
right diagonals), and total absorption components (pure + P-Cygni
+ inverse P-Cygni). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.7 v50 and v84 as a function of the stellar masses. The meanings of
the symbols are the same as those in Section 2.5.1.1. The data
points joined by a segment correspond to F14394+5332 W and E.
The smaller symbols have larger uncertainties (values followed by
double colons in Table 3.3) . . . . . . . . . . . . . . . . . . . . . . . . 67
2.8 v50 and v84 as a function of the star formation rates. The meanings
of the symbols are the same as those in Section 2.5.1.1. The smaller
symbols have larger uncertainties (values followed by double colons
in Table 3.3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.9 v50 and v84 as a function of the specific star formation rates. The
meanings of the symbols are the same as those in Section 2.5.1.1.
The data points joined by a segment correspond to F14394+5332 E
and W. The smaller symbols have larger uncertainties (values followed
by double colons in Table 3.3) The black vertical line indicates the
approximate location of the Main Sequence of star-forming galaxies
(Shimizu et al., 2015). . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.10 v50 and v84 as a function of the AGN fractions. The meanings of the
symbols are the same as those in Section 2.5.1.1. The smaller symbols
have larger uncertainties (values followed by double colons in Table 3.3) 70
2.11 v50 and v84 as a function of the AGN luminosities. The meanings of
the symbols are the same as those in Section 2.5.1.1. The smaller
symbols have larger uncertainties (values followed by double colons
in Table 3.3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
x
2.12 Total (absorption + emission) equivalent widths of OH 119 ?m as
a function of the apparent strength of the 9.7 ?m silicate feature
relative to the local mid-infrared continuum. Note that S9.7?m is a
logarithmic quantity and can be interpreted as the apparent silicate
optical depth. The strength of this absorption feature increases to the
right. Also note that objects classified as LINERs have been excluded
from this plot. Filled markers refer to Type 1 and open markers refer
to Type 2. Blue squares and red circles represent BAT AGN and
ULIRGs/PG QSOs, respectively. Vertical lines represent objects in
which OH was not detected. Horizontal lines represent objects with
a null S9.7?m detection. Dotted lines and dash-dotted lines refer to
Type 1 and Type 2, respectively. Blue and red lines indicate BAT
AGN and ULIRGs/PG QSOs, respectively. . . . . . . . . . . . . . . . 80
2.13 Ratio of the semi-minor axis to the semi-major axis (a proxy for the
inclination of the host galaxy disk) as a function of S9.7?m for the BAT
AGN sample. Squares, triangles, and circles represent BAT AGN in
which OH is observed purely in absorption, purely in emission, com-
posite absorption/emission, respectively. Diamonds represent objects
in which OH was undetected. Filled points and dash-dotted lines in-
dicate Type 1 while open points and dotted lines indicate Type 2
AGN. Horizontal lines represent objects with a null 9.7 ?m silicate
feature detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.1 Example of twelve different continuum fits to the mid-infrared spec-
trum of IRAS F17208-0014. The black solid histograms are the data
separated from each other by an arbitrary offset. Spline pivot points
are anchored at wavelengths of 33.0 ?m, 33.8 ?m, 34.5 ?m, and 35.0
?m. The blue, green, and red dots indicate the set of pivot points
which are respectively the lower limits (v1), upper limits (v2), and
best-fit-by-eye values (v3) used to fit a continuum spline. The top
most set of fits is the result of using a third-order (k = 3) spline fit to
each pivot point set. Below are the results for second-order (k = 2)
and first-order (k = 1) spline fits. In each case, the blue, green, and
red dotted lines show the results of the spline fits to the lower-limit,
upper-limit, and ?best-fit-by-eye? continuum flux pivot points. The
last spectrum at the bottom is fit with a polynomial. The magenta
regions in this spectrum indicate the regions used to fit the first (solid
blue line), second (solid green line), and third (solid red line) order
polynomials to the continuum. The two dotted vertical lines in grey
mark the rest wavelengths of the OH 35 doublet at systemic velocity.
The dotted vertical lines in red mark the locations of the emission fea-
tures [S III] 33.48 ?m and [Si II] 34.82 ?m. The vertical light-green
band shows the region where a 20% dip in the detection response in
the NOD 1 position occurs. . . . . . . . . . . . . . . . . . . . . . . . 94
xi
3.2 Fits to the 33? 35?m continua for all 15 objects in the sample. The
grey shaded area shows the full range of the twelve different contin-
uum fits described in Figure 3.1. Black dots are median values of
the six fluxes at each pivot point, and the dotted line is the resultant
third-order spline fit to those pivot points. The grey dotted vertical
lines mark the rest wavelengths of the OH 35 doublet at systemic
velocity. The red dotted vertical lines mark the locations of the emis-
sion features [S III] at 33.48 ?m and [Si II] at 34.815 ?m. The vertical
light-green band shows the region where a 20% dip in the detection
response in the NOD 1 position occurs. . . . . . . . . . . . . . . . . . 95
3.3 Two-Gaussian fits to the continuum-subtracted OH 35 line profiles
of the 15 objects in our sample; see Section 3.3.3. In each figure, the
solid black histogram is the data. Blue dashed lines indicate the two
Gaussian components which best fit the line profile, and the magenta
line is the sum of those two components. The grey dotted vertical
lines mark the rest wavelengths of the OH 35 doublet at systemic
velocity. The red dotted vertical line marks the location of the [Si II]
emission line at 34.815 ?m. . . . . . . . . . . . . . . . . . . . . . . . 98
3.4 Examples of OH 35 line profile fits for each of the twelve different
continuum subtracted spectra in IRAS F17208-0014. Line colors and
styles are the same as that for Figure 3.3. ?p1, p2, p3? indicate first-,
second-, and third-order polynomial fits to the continuum, respec-
tively. ?v1? refers to the ?lower-limit? pivot points in Figure 3.1.
?v2? refers to the ?upper-limit? pivot points, and ?v3? to the ?best-
fit-by-eye? pivot points. ?k1, k2, k3? are the orders of the spline fitted
to the continuum. For example, ?v2k3? is the third-order spline fit
to the continuum using the ?upper-limit? pivot points. . . . . . . . . 99
3.5 Distributions of the measured OH 35 (a) integrated fluxes and (b)
equivalent widths. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
3.6 Distributions of the (a) OH column densities and (b) OH-to-H abun-
dance ratios, XOH, inferred from the fits of the OH 35 feature. . . . . 102
3.7 OH 35 spectra normalized to the continuum. Red and blue lines are
the two Gaussian components of the fitted line and the magenta line
is the sum of the two components. The green line is the profile pre-
dicted by the radiative transfer model described in Gonza?lez-Alfonso
& Cernicharo (1999); Gonza?lez-Alfonso et al. (2017). The modeled
line profiles are constrained by observations of four OH lines at 65,
79, 84, and 119 ?m. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.1 Signal-to-noise ratio maps. Contours are Contours are 0.1, 0.3, 0.5,
0.7, 0.8, and 0.9 of the peak value in each image. Black crosses mark
the adopted galaxy center. . . . . . . . . . . . . . . . . . . . . . . . . 124
4.2 Total integrated emission line fluxes in 10?17 W m?2. Contours are
0.1, 0.3, 0.5, 0.7, and 0.9 of the peak flux in the image. Black crosses
mark the adopted galaxy center. . . . . . . . . . . . . . . . . . . . . . 125
xii
4.3 Maps of the emission line ratios. Contours are 0.1, 0.3, 0.5, 0.7, 0.8,
and 0.9 of the peak value in each image. . . . . . . . . . . . . . . . . 131
4.4 Maps of the median velocities, v in units of km s?150 . Contours are
in eight equal steps between the minimum velocity and the maximum
velocity in each image. . . . . . . . . . . . . . . . . . . . . . . . . . . 132
4.5 Maps of the 1?? line widths, W1? , in units of km s?1 . Contours are
0.4, 0.5, 0.6, 0.7, 0.8, and 0.9 of the peak width in each image. . . . . 133
4.6 Maps of the hydrogen density, nH in each object. Units are in cm
?3.
The ?low? density range is defined as 0 < nH< 10
4 cm?3 and the
?high? density range is defined as 104 < nH< 10
7 cm?3. See Sec-
tion 4.7.4 for details about the determination of these density ranges
and definitions of Solutions A?D. Solution A is the low density
solution with only the [C II] 158 flux correction (see Section 4.7.1
and Section 4.7.4), Solution B is the low density solution with
both the fluxes of [C II] 158 and [O I] 63 corrected (see Section 4.7.1,
Section 4.7.2 and Section 4.7.4), Solution C is the high density
solution with only the [C II] 158 flux correction, and Solution D
is the high density solution with both the fluxes of [C II] 158 and
[O I] 63 corrected. As discussed in Section 4.7.4, the high density
PDR solutions ?C? and ?D? lead to unphysical ISRFs. Therefore, we
have excluded them from the analysis of the outflow. . . . . . . . . . 139
4.7 Maps of the strength of the UV radiation field, G0, for the low density
and high density limits PDR solutions. Definitions of the Solutions
are the same as those in Figure 4.6. . . . . . . . . . . . . . . . . . . . 140
4.8 M 82: Results from modeling the disk velocity field with 3DBarolo (which
accounts for both the instrumental spectral and spatial resolutions)
and the location of the outflow based on excess line broadening. Color
bar values are in units of km s?1 . For each line, left to right: observed
data, model result, data - model residual, spatial location of the wind
in regions where ?W > 25 km s?11? , and spatial location of the wind
in regions where ?W1? > 50 km s
?1 . Contours in (a) are in five equal
steps between the minimum and maximum velocities in each image.
Contours in (b) are 0.3, 0.5, 0.7, and 0.9 of the peak value in each
image. The solid blue line marks the galaxy major axis. The magenta
cross marks the adopted galaxy center. . . . . . . . . . . . . . . . . . 156
4.9 M 82: PACS contours (magenta) overlaid on the total integrated
intensity contours of SiO(2-1) (black, Garc??a-Burillo et al., 2001)
and the radio continuum at 4.8 GHz (grey scale, Wills et al., 1999). . 157
4.10 Mass and KE in the wind derived from the PDR solutions (see Sec-
tion 4.7.4 for definitions of Solutions A, B, C, and D) and the [O III] 88 flux.
Units are in 104M and 1051 erg for the masses and KEs, respectively.158
4.11 Cen A: Results from modeling the disk velocity field with 3DBarolo.
Symbols and plots are the same as those in Figure 4.8. . . . . . . . . 161
4.12 Cen A: PACS contours overlaid on 8.4 GHz image from Hardcastle
et al. (2003). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
xiii
4.13 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in
Figure 4.10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
4.14 Circinus: Results from modeling the disk velocity field with 3DBarolo.
Symbols, units, and plots are the same as those in Figure 4.8. . . . . 165
4.15 Circinus does not have a discernible outflow in either the neutral or
ionized gas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
4.16 NGC 253: Results from modeling the disk velocity field with 3DBarolo.
Symbols, units, and plots are the same as those in Figure 4.8. . . . . 168
4.17 NGC 253: PACS contours (yellow), contours of receding and ap-
proaching CO outflow (magenta and blue, respectively) from Bolatto
et al. (2013). Composite image from Heesen et al. (2011) which
shows H? in red from Westmoquette et al. (2011), ?20 cm contin-
uum (green), and soft X-ray in blue from Hardcastle et al. (2010).
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
4.18 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in
Figure 4.10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
4.19 NGC 1068: Results from modeling the disk velocity field with 3DBarolo.
Symbols, units, and plots are the same as those in Figure 4.8. . . . . 172
4.20 NGC 1068: PACS contours overlaid on 349 GHz continuum and
CO(3-2) residual mean-velocity field from Garc??a-Burillo et al. (2014). 173
4.21 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in
Figure 4.10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
4.22 NGC 3079: Results from modeling the disk velocity field with 3DBarolo.
Symbols, units, and plots are the same as those in Figure 4.8. . . . . 177
4.23 Top row: PACS contours of the [C II] 158 v50 and W1? residuals in the
wind in NGC 3079 overlaid on 1.4GHz observations from Sebastian
et al. (2019). Middle row: PACS contours of the [C II] 158 v50 and
W1? residuals in the wind in NGC 3079 overlaid on H?+ [N II] image
from Cecil et al. (2001). Bottom row: Chandra image (blue) + HST
(red and green). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
4.24 NGC 4945: Results from modeling the disk velocity field with 3DBarolo.
Symbols, units, and plots are the same as those in Figure 4.8. . . . . 180
4.25 Top row: PACS contours overlaid on the [NII] residual velocity field
of NGC 4945 from Venturi et al. (2017). Bottom row: PACS contours
overlaid on the [NII] W70 of NGC 4945 from Venturi et al. (2017). . . 181
4.26 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in
Figure 4.10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
xiv
A.1 Mid-infrared (5-37?m) spectra used to measure S9.7?m . The dashed
line is the continuum calculated from the cubic spline interpolation fit-
ted to the pivot points shown as black dots. Red dots show fcont(9.7?m)
(located on dashed continuum line) and fobs(9.7?m) (located on the
solid black line or the observed flux density). The blue line shows
the integration range used to calculate the flux and total equivalent
width of the 9.7 ?m silicate feature (see Table 2.4 and Table 2.5). . . 192
A.2 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
A.3 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
A.4 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
A.5 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
A.6 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
A.7 (Continued) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
B.1 Top left: PACS IFU footprint of OH 119 (black squares) overlaid
on the 22 ?m WISE image of M 82. The white contour marks the
spatial extent of the [O III] 88 wind. The black contour outlines the
PACS footprint of the [O III] 88 observation. The black cross marks
the adopted galaxy center and the blue line marks the galaxy major
axis. Top right: Spline fits to the OH 119 continuum (blue dashed
lines). Black lines are the observed data. Magenta areas indicate the
regions used to fit the continuum. Blue dots mark the pivot points
used to fit the spline. Bottom left: Line profile fitting results of
the continuum-subtracted spectra. Solid blue lines indicate gaussian
absorption components. Solid red lines indicate gaussian emission
components. Vertical dashed blue (red) lines mark the v16 , v50 , and
v84 velocities in absorption (emission). Bottom right: Top row, from
left to right shows the total velocity-integrated flux of the fitted OH
line profiles, the total flux in the absorption components only, and the
total flux in the emission components only. Middle row shows v50 for
the absorption (left) and emission components (right). Bottom row
shows the 1?? line widths of the absorption (left) and emission (right)
components. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
B.2 Cen A. Lines and symbols are the same as those in Figure B.1. . . . . 203
B.3 Circinus. Lines and symbols are the same as those in Figure B.1. . . . 204
B.4 NGC 253. Lines and symbols are the same as those in Figure B.1. . . 205
B.5 NGC 1068. Lines and symbols are the same as those in Figure B.1. . 206
B.6 NGC 3079. Top left: The white contour marks the spatial extent
of the [C II] 158 wind. Lines and symbols are the same as those in
Figure B.1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
B.7 NGC 4945. Lines and symbols are the same as those in Figure B.1. . 208
xv
List of Abbreviations
AGN Active Galactic Nucleus
ALMA Atacama Large Millimeter Array
BAT Burst Alert Telescope
CASSIS Cornell AtlaS of Spiter/IRS Sources
EW Equivalent Width
FIR Far-Infrared
HST Hubble Space Telescope
IR Infrared
ICM Intracluster Medium
IGM Intergalactic Medium
IRS Infrared Spectrograph
ISM Interstellar Medium
IFS Integral Field Spectroscopy
IFU Integral Field Unit
LF Luminosity Function
LIRG Luminous Infrared Galaxy
MIR Mid-Infrared
NED NASA/IPAC Extragalactic Database
PACS Photodetector Array Camera and Spectrometer
PAH Polycyclic Aromatic Hydrocarbon
PG QSO Palomar Green Quasi-Stellar Object
RSRF Relative Spectral Response Function
SB Starburst
SED Spectral Energy Distribution
SHA Spitzer Heritage Archive
SMBH Supermassive Black Hole
SF Star Formation
SNe Supernovae
sSFR Specific Star Formation Rate
SFR Star Formation Rate
ULIRG Ultra-Luminous Infrared Galaxy
UV Ultraviolet
xvi
Chapter 1: Introduction
1.1 Background
Galactic-scale outflows (or winds) are powered by stellar and/or supermassive
black-hole (SMBH) accretion processes which inject energy and momentum into
the interstellar medium (ISM). On a grand scale, winds may significantly affect the
assembly and evolution of galaxies (Veilleux et al., 2005, 2020). Outflows are ubiq-
uitous at any redshift and they are invoked to help explain a number of observations
in the Universe. For example, they help explain some of the observable properties
of galaxies (e.g. color and gas abundance; for example, red, gas-poor ellipticals) and
of the intergalactic medium (IGM). Outflows can account for the correlations that
exist between the physical properties of a galaxy (e.g. the tight correlation between
the SMBH mass and the mass of the spheroidal bulge). Moreover, outflows may
help reconcile (or mitigate) the discrepancies between predictions from cosmological
simulations and models with actual observations. In the following list, I expound
upon some specific scenarios where winds have been incorporated to explain a num-
ber of observations. This list is by no means exhaustive, but it does highlight how
winds may provide a significant contribution to our understanding of the formation
and evolution of galaxies.
1
? Star formation regulation - There are a number of ways an outflow can affect
the star formation activity in a galaxy. If the outflow has enough energy to
escape the gravitational potential well of its host galaxy, then the outflow may
be able to empty the reservoir of gas within the system and halt star formation
completely (Fabian, 2012; Veilleux et al., 2005). Winds may heat the cold ISM
via shocks or from the deposition of mechanical energy, thereby preventing
the gas from condensing and forming stars (e.g. Cecil et al., 2001; Heckman
& Thompson, 2017; Martin, 2006). On the other hand, if the material in the
outflow lacks the energy to escape its host system, it will fall back onto the
disk in the form of a ?galactic fountain? and may thus be recycled again for
star formation.
? Scaling relations with the SMBH mass and the host galaxy properties - Winds
may be responsible for the fueling or quenching of accretion onto the SMBH in
a galaxy (Di Matteo et al., 2005). The brightness of an active galactic nucleus
(AGN) is regulated by the rate of infall from surrounding material (e.g. gas and
dust) while the infall rate of material is regulated by the brightness of the AGN.
This feedback loop may explain the remarkable correlations observed between
the mass of the SMBH (MBH) and the properties of the host galaxy, such
as luminosity (MBH?L relation), velocity dispersions in (MBH?? relation)
and masses (MBH?Mbulge relation) of the spheroidal bulge (see the review and
references therein of Kormendy & Ho, 2013). These relationships indicate that
SMBHs and the bulges coevolve by regulating each other?s growth. Figure 1.1
2
shows the three relationships.
? Correlations with galaxy type - Winds that can empty their host galaxies of gas
may explain the bimodal color distribution between early (E, S0, and Sa) and
late (Sb, Sc, and Irr) morphological types observed in large galaxy surveys (e.g.
Strateva et al., 2001). These surveys indicate that disk-dominated galaxies are
predominantly blue and star forming, whereas spheroid-dominated galaxies are
largely red and quiescent (e.g. Blanton & Moustakas, 2009). This bimodality
indicated that some sort of quenching mechanism was required to prevent gas
from cooling and/or forming stars in observed gas-poor, quiescent galaxies
(Peng et al., 2010). The inclusion of winds into semi-analytic models has
qualitatively reproduced this observed bimodal color distribution (e.g. Kimm
et al., 2009).
? Metal enrichment of the IGM and CGM - Outflows may not only impact in-
dividual galaxies, but they may also play a significant role in the chemical
and thermal evolution of the universe. They may even influence the metal en-
richment of the intergalactic medium (IGM) and the circumgalactic medium
(CGM) via the regulation of gas and dust exchange into and out of galaxies
(Bertone et al., 2007; Heckman & Thompson, 2017; Zhang, 2018). Observa-
tions of absorption systems in quasar sightlines through the IGM/CGM show
that the IGM/CGM is enriched with metals to a small fraction of solar metal-
licity and simulations have shown how outflows may be responsible for lifting
gas and metal up and out of the galaxy disk into the IGM/CGM (Aguirre
3
et al., 2001; Madau et al., 2001).
? Paucity of very high and low mass galaxies - Early attempts to predict the
statistical distributions of global properties of galaxy populations have shown
a mismatch in shape between the predicted distribution of dark halo masses
and the observed distribution of galaxy luminosities (or equivalently, masses).
Galaxy abundances at higher luminosities (e.g. those brighter than galaxies
with Milky Way-like luminosities, L? ? 3 ? 1010 L) are observed to decline
exponentially compared to CDM simulations, while the shape of galaxy abun-
dances are observed to flatten at the fainter end. Inclusion of feedback from
the SMBH appears to explain the abundance deficit at the high luminosity
end (e.g. Croton et al., 2006; Somerville et al., 2008), while the inclusion of
supernovae (SNe) and stellar processes is able to reproduce the shape of the
luminosity function (LF) at the fainter end (e.g. Bower et al., 2006). Essen-
tially, feedback due to stellar processes plays a crucial role in shaping the LF
for dark matter halos with MH . 1012Mwhile feedback from AGN processes
plays a more dominant role at halo masses M & 1012H M (Somerville & Dave?,
2015).
Although we now have a better understanding of how winds may significantly
impact the formation and evolution of galaxies there still exists a plethora of unan-
swered questions concerning the detailed physics and properties of galactic outflows.
What is the primary mechanism of energy and momentum injection that powers an
outflow? How can we constrain outflow properties such as energetics, morphologies,
4
b
Figure 1.1 Left and Middle: MBH?L and MBH?? relations, respectively (from
Gu?ltekin et al., 2009). Right: MBH?Mbulge relation (from Kormendy & Ho, 2013).
lifetimes, spatial extent, radial-dependent velocities, mass outflow rates? What is
the efficiency of entraining material from the ISM and ejecting it out of the disk?
The answers to these questions can fundamentally measure the impact of outflows
on galaxy evolution.
In this chapter, I describe the basic physics behind outflows. In particular, the
sources of energy and momentum (either starburst (SB) or AGN) and the driving
mechanisms of the outflows. I also discuss the multiphase nature of outflows and
the tools we used to probe different ISM phases in outflows.
1.2 Starburst and Active Galaxies
1.2.1 Starburst-Driven Winds
SB galaxies contain regions (typically nuclear or circumnuclear) that are un-
dergoing intense and spatially concentrated star formation activity; much higher
than the rate of star formation in galaxies similar to the Milky Way. Such a SB can
5
be triggered when gas rapidly accumulates into a small volume, say from a merger
of two gas-rich galaxies Barnes & Hernquist (1996), or from gas funneled into the
nuclear region of a galaxy by dynamical processes associated with bar instabilities
Combes & Charmandaris (2000). The SB regions host large numbers of young, hot,
massive stars that ionize their environment and evolve quickly to form supernovae.
Outflows can be driven by the powerful energy reservoirs in these SB regions.
Chevalier & Clegg (1985, hereafter, C85) was first to establish a conceptual
picture of how a SB-driven wind would work. The evolution of a bubble in the in-
terstellar medium (ISM) evolves in several stages. First, kinetic energy is launched
from a collection of stellar processes such as winds from high-mass, high-luminosity
stars and explosions from supernovae (SNe). The ejecta from the winds and explo-
sions intersect and collide with each other creating shocks that rapidly thermalize
the kinetic energy of the ejecta, thus producing hot (T ? 107 K), high pressure
(P/k ? 107 K cm?3), low density gas (Strickland & Stevens, 2000). The wind-
shocked material creates an over-pressurized cavity (or bubble) that adiabatically
expands as a free wind with vwind ? several thousand km s?1 (Garc??a-Burillo et al.,
2001). During this ?energy-conserving? phase, little energy is lost to radiation due
to the high post-shock temperatures and low densities. In the next phase, the free
wind passes through an internal wind shock where the ejecta is decelerated and re-
heated to T ? 107 K. The pressure of the shocked wind continues to inflate a bubble
which is collimated by the density gradient between the galactic disk and its halo.
The bubble preferentially propagates along the direction of the steepest pressure
gradient. This forms a bipolar flow which sweeps up cooler, denser, ambient disk or
6
halo gas that is accelerated to vgas ? few hundreds km s?1 . A shell of accumulated
ISM gas then forms at the edge of the bubble. Once the velocity of the shell has
decelerated enough so that the cooling time is shorter than the expansion time, the
shell will begin to cool radiatively. The bubble then enters the pressure-driven snow-
plow phase in which the internal thermal energy of the bubble cannot be neglected.
The bubble leaves the snowplow phase once the thermal energy has been radiated
away. The total momentum of the inner wind is transferred to the gas and the
bubble enters a ?momentum-conserving? phase. Assuming a stratified halo, once
the bubble diameter is comparable with the scale height of the disk, the ambient gas
density will be low enough that the bubble can accelerate and fragment into clumps
and filaments through Rayleigh-Taylor instabilities (Schiano, 1985). During this
?break-out? phase, the freely flowing and shocked wind is injected into the galaxy
halo.
In summary, the structure of a bubble expanding into an ISM due to a SB
includes (a) the energy injection zone, (b) hot, freely-expanding wind, (c) wind ma-
terial that has been shocked, (d) a thin, dense shell of post-shock gas that has been
accumulated by the wind and is now radiatively cooling, and finally (e) undisturbed
ISM material. A schematic displaying the essential features of a wind-blown bubble
is shown in Figure 1.2.
7
Figure 1.2 A schematic showing two evolutionary phases of a wind-blown bubble.
Panel (a) shows the radiative phase where the energy injection zone from the SB is
indicated by open stars, the free wind (FW, long arrows) expands outward at several
thousand km s?1 where it passes through an internal wind shock (WS) creating
shocked wind material (SW). The expanding bubble will sweep up ambient gas and
accelerate it to a few hundred km s?1 (short arrows). The shocked gas will then
radiatively cool into a thin shell (S). Eventually, when the diameter of the bubble
is comparable to the scale height of the disk, it will fragment, panel (b), where the
freely expanding wind fluid will flow into the halo. Bow shocks (BS) can form around
shocked clouds (SC) which may be accelerated outward (from Heckman et al., 1990).
1.2.2 AGN-Driven Winds
An active galaxy contains a core of bright emission that is embedded in the
nuclear region of an otherwise normal galaxy. This core is typically brighter than the
rest of the galaxy, shows variations in luminosity, and is known as an AGN. Spectral
energy distributions (SEDs) of this higher-than-normal luminosity core indicate that
the excess brightness is non-stellar in origin. It is generally accepted that the source
of energy from an AGN is the accretion of matter (e.g. gas or dust) onto the SMBH,
where the accreted material loses its gravitational potential energy via conversion
into heat and is partly radiated away. This accretion disk is mostly seen in the
8
ultra-violet (UV) and in X-rays.
In AGN, the nature of the energy outflow near the SMBH differentiates AGN
feedback into two modes:?radiative? or ?quasar? mode which is powered by radiation
and ?radio? or ?kinetic? mode which is powered by mechanical jets (Cielo et al.,
2018). Radiative mode feedback is typically observed in high-luminosity AGN that
are radiatively efficient and close to the Eddington limit (within about one or two
orders of magnitude). It is most concerned with pushing cold gas about (Fabian
et al., 2013). This mode is the most likely explanation for the observed black hole
mass?stellar velocity dispersion relation (the so-called ?MBH??? relation). This
mode was probably most effective at z ? 2 ? 3 when quasar activity was more
common and galaxies were more rich in gas. The radio mode AGN are radiatively
inefficient and typically host the most massive SMBH. These low-luminosity AGN
(LAGN < 0.01LEdd; Combes, 2014) typically harbor light relativistic radio jets
and are surrounded by hot halos. They typically correspond to massive early type
galaxies (e.g. radio ellipticals). If feedback successfully empties a galaxy of its gas,
eventually stellar mass loss and/or intracluster plasma will inevitably refill the gas
reservoir. Maintaining an empty reservoir, or at least heating the gas enough to
prevent star formation, appears to be a consequence of the radio mode feedback.
In general, an AGN-driven wind will collide with the surrounding ISM, driving
a reverse shock into the wind and a forward shock into the ISM. If the shocked wind
cools efficiently, the outflow is ?momentum-conserving? (or ?momentum-driven?)
and transfers only its ram pressure (and, hence, momentum flux) to the ISM (Zubo-
vas & Nayakshin, 2014). If, on the other hand, the shocked wind cools inefficiently,
9
Figure 1.3 A schematic view of the shock driven into the ISM by a wind launched
from an AGN is displayed in the left image where (a) is the wind region, delimited by
the reverse wind shock (RSW ), (b) is the shocked wind that ends at the discontinuity
surface (RC), and (c) is the shocked ISM, bounded by the forward shock (RS). The
top right image shows a schematic of the ?energy-conserving? outflow where the hot
and thick region (b) cools inefficiently and expands adiabatically. The bottom right
image shows the ?momentum-conserving? outflow where the shocked wind (light
blue region (b)) cools efficiently, experiences a drop in pressure, and becomes thin
(from Costa et al., 2014).
the outflow is ?energy-conserving? (or ?energy-driven?) and transfers most of its
kinetic luminosity (Lkin ' 0.05LAGN; Zubovas & Nayakshin, 2014) to the ambient
medium. This hot outflow may create a bubble in the galaxy disk that thermally
expands after the shock. This phenomenon is analogous to the adiabatic phase in a
SB-driven bubble described in Section 1.2.1. This kind of wind can drive outflows
to velocities greater than 1000 km s?1 (King et al., 2011). A schematic view of the
shock driven into the ISM by a wind launched from an AGN is displayed in the
left image of Figure 1.3. The two images on the right of Figure 1.3 show the two
10
different driving modes.
The bubble scenario described above is valid for an AGN driving a wide-
angle outflow from the accretion disk or for a low energy AGN jet where all of the
kinetic energy of the jet is deposited to the ISM. If the jet is powerful and highly
collimated, the jet will simply ?drill? through the ISM without imparting much
energy or momentum to the surrounding gas (e.g. Scheuer, 1974).
1.3 The Multiphase ISM in Outflows
1.3.1 General Considerations
It is well established that outflows are of a multiphase nature (as revealed
in observations and expected from simulations), spanning from the cold and dense
molecular clouds to the very hot highly ionized medium. This multiphase structure
complicates the calculations for the masses and energetics in outflows because we
must concurrently analyze the outflow in multiple wavelength regimes, otherwise,
measurements based on a single-phase will most likely result in misleading and
incomplete conclusions. However, if we wish to assess the degree to which galactic
winds impact the evolution and formation of galaxies, a fundamental understanding
of the dynamic, energetic, and physical properties of the outflow must be established.
Therefore, it is essential to consider all phases of the ISM in outflows in order to
gain an understanding of the cosmological significance of outflows.
11
Figure 1.4 Artistic and observational views of AGN-driven winds in different ISM
phases and at different physical scales. The cartoon in panels a, b, and c depicts
the multiphase outflow based on observations and AGN feedback models. The
outflow is launched from the galaxy nucleus (< 1 pc; panel a), propagates through
the surrounding host galaxy ISM (1 pc ? 1 kpc; panel b), and extends out into
the galaxy halo (> 10 kpc; panel c). Observations of different outflow phases at
different physical scales in three well known AGN are shown in panels d, e, and
f. Panel d shows the highly ionized accretion disk wind observed in X-ray from
PDS456 (Nardini et al., 2015). Panel e shows the neutral atomic and molecular
outflow phases traced by the mm and radio in Mrk 231 on a scale of ? few hundreds
pc (Cicone et al., 2012; Morganti et al., 2016). Finally, panel f shows the ionized
optical outflow of NGC 1365 extending out a few kpc (Venturi et al., 2017). (Figure
is reproduced from (Cicone et al., 2018)).
12
Multiwavelength observations from the radio to the ??ray have shown that
the multiphase structure of an outflow are generally divided into four gas phases: hot
(highly ionized, Tgas? 106?107 K; n 6 8 ?3gas? 10 ?10 cm ), warm (ionized, T ? 103gas ?
104 K; n 2 4 ?3gas? 10 ?10 cm ), cool (neutral atomic, T 2 3gas? 10 ?10 K; ngas? 1?102
cm?3), and cold (molecular, Tgas? 10?102 K; n ? 103 cm?3gas ). These phases have
been studied in several objects through various techniques which span the length
of the electromagnetic spectrum. I will elucidate how these ISM phases have been
traced in previous studies and I note that this list is by no means exhaustive or
complete.
Chandra and XMM-Newton have been used to observe the highly ionized X-
ray emitting gas (e.g. Bravo-Guerrero & Stevens, 2017; Strickland & Heckman, 2007;
Strickland & Stevens, 2000). Optical emission (e.g. H? , [N II], [O III]) and absorp-
tion (e.g. Na I D and Mg II) lines have been used to trace the warm ionized (Cecil
et al., 2002; Heckman et al., 2015; Martin et al., 2013; Rubin et al., 2014; Shop-
bell & Bland-Hawthorn, 1998; Weiner et al., 2009) and neutral atomic (Heckman
et al., 2000; Kornei et al., 2013; Martin, 2005; Rupke et al., 2002, 2005a,b,c; Rupke
& Veilleux, 2015) gas phases in outflows. The warm molecular phase in outflows
has been traced with H2 (e.g. Veilleux et al., 2009). Cold molecular gas has been
traced using OH, CO, HCN, and HCO+ (e.g. Bolatto et al., 2013; Leroy et al., 2015;
Veilleux et al., 2013; Walter et al., 2017; Westmoquette et al., 2013). Dust has also
been traced in outflows (e.g. Hutton et al., 2014; McCormick et al., 2018; Mele?ndez
et al., 2015). Figure 1.4 shows artistic representations and observational data of dif-
ferent phases of AGN-driven outflows at different physical scales. Figure 1.5 shows
13
Figure 1.5 A schematic of the multiphase outflow in the SB galaxy M 82. The
red statements are supported by the observational evidence described in blue (from
Leroy et al., 2015).
the multiphase SB-driven outflow in M 82.
1.3.2 The OH Molecule
We use the FIR lines of the hydroxyl molecule (OH) as a tracer of the molecular
gas component in an outflow. Furthermore, these lines can be used as powerful
diagnostics of molecular winds because of their operative excitation mechanisms
and their range in optical depths. These properties (e.g. OH is mainly excited by
the absorption of FIR) also allow OH to probe nuclear regions and star forming
complexes within galaxies where the FIR radiation density is strong. Within the 34
14
?m to 163 ?m FIR wavelength range, a total of 14 lines (exhibiting both optically
thick and optically thin features) arise from the eight lowest rotational levels of OH.
In this thesis, we focus on two of the OH transitions. The first is the ground-
state OH 2?3/2 J = 5/2? 3/2 rotational ?-doublet at 119.233 ?m and 119.441 ?m
(hereafter, OH 119). It is the strongest of these FIR OH transitions (therefore it is
more likely to be detected) and its wavelength is conveniently positioned at the peak
sensitivity range of Herschel -PACS. Previous studies have already demonstrated
the powerful diagnostic capability of OH 119 to determine wind characteristics (e.g.
Fischer et al., 2010; Spoon et al., 2013; Sturm et al., 2011; Veilleux et al., 2013).
Moreover, these studies have shown that the molecular gas can dominate the mass
and energy budget of the galactic outflow.
We also look at the cross ladder OH 2? ?23/2 ?5/2 ??doublet transition at
34.60 and 34.63?m (hereafter, OH 35). OH 35 is optically thin compared to the
other FIR OH transitions and because only 4% of all 35?m absorption events result
in re-emission at 35?m, OH 35 absorption line can be used to provide a meaningful
constraint on the true column density.
See Figure 1.6 for a schematic of the OH energy levels. Transitions analyzed
in this thesis are outlined in red.
1.3.3 Atomic Fine-Structure Lines
Fine-structure line emission from neutral and ionized carbon atoms, as well as
other species such as oxygen and nitrogen, is an important key to understanding the
15
Figure 1.6 OH ?-doublet transitions. Transitions analyzed in this thesis are outlined
in red.
properties of the ISM in the region from which they originate, as it provides almost
all of the gas cooling.
Warm, ionized, and relatively tenuous gas can be traced by [O III] 88, [N II] 122,
and (to a smaller extent) [C II] 158. Warm, neutral, dense gas is exclusively traced
by [O I] 63 and [O I] 145, but [C II] 158 emission also predominately occurs in the
atomic medium. The observed lines cover a range in density and temperature be-
havior, which probe different phases of the ISM and are expected to probe different
regions of their host galaxies. Figure 1.7 shows a schematic of the energy levels for
the five atomic fine structure lines.
16
Figure 1.7 Energy level diagrams of the transitions in the electronic ground states of
the atomic fine-structure lines. Top left: C+ (Draine, 2011). Top right: Neutral O
(Draine, 2011). Bottom left: N+ ion (Goldsmith et al., 2015). Bottom right: O++
ion. The optical transitions are on the left and the fine-structure transitions are on
an expanded scale on the right (Dinerstein et al., 1985).
1.4 Observatories and Instruments
The online archival data presented in this thesis were acquired with the Her-
schel Space Observatory and the Spitzer Space Telescope. Both telescopes have run
through their cryogenic fuel and are no longer in operation. In this section, I detail
these observatories and the instruments used to obtain the data analyzed in this
thesis. Chapter 2 utilizes PACS and Spitzer. Chapter 3 utilizes only Spitzer data.
Chapter 4 utilizes only PACS data.
17
Figure 1.8 Top left: Herschel Space Observatory that carried PACS. Top right:
Spitzer Space Telescope carried IRS. Bottom: Image shows a series of model spec-
tral energy distributions (SEDs) for a star-forming galaxy. Wavelength coverage
is indicated for other IR instruments. PACS covers the principal FIR atomic fine
structure cooling lines of the ISM: [O I] 63 , [O III] 88 , [N II] 122 , [O I] 145 , and
[C II] 158 . Notice that OH 119 falls at the peak sensitivity of PACS.
1.4.1 Herschel -PACS
The bulk of the data in this thesis are FIR observations from the Herschel
Space Observatory (Pilbratt et al., 2010) which carries a 3.5 m diameter Cassegrain
18
telescope. It is the largest infrared space telescope to have ever been launched.
Spectroscopic data was taken with the Photodetector Array Camera and Spectrom-
eter (PACS; Poglitsch et al., 2010) on board the Herschel. PACS covers a field
of view of 47??? 47??with a 5 ? 5 squared spaxel (spatial pixel) integral field unit
spectrograph. It has a wavelength dependent spectral resolving power that ranges
from R = 1000 ? 4000 (?v ? 75 ? 300 km s?1 ). PACS sampled the ? 50 ? 210
?m FIR wavelength range with an unprecedented combination of sensitivity and
spatial resolution (? 6 ? 11??; this is about 4? higher than previous infrared space
missions). PACS covers the principal FIR fine-structure cooling lines (see bottom
panel of Figure 1.8). Archival Herschel -PACS data were retrieved via the Herschel
Science Archive (HSA).
1.4.2 Spitzer -IRS
Launched in 2003 The Spitzer Space Telescope (Werner et al., 2004), operated
by the Jet Propulsion Laboratory and the Spitzer Science Center (SSC), incorporates
a 0.85 m diameter primary mirror and has a wavelength coverage from 3.6 to 160?m.
The mid-infrared (5 ? 37?m), ?high-resolution? (R ? 600) data presented in this
thesis were obtained with the Long High (LH) module of the Infrared Spectrograph
(IRS; Houck et al., 2004). The data were retrieved from The Cornell AtlaS of
Spitzer/IRS Sources (CASSIS) or from the Spitzer Heritage Archive (SHA).
19
1.5 Outline of Thesis
This thesis focuses on the study of cool galactic winds in SBs and AGNs.
Chapters 2 and 3 focus on the molecular phase in outflows. Chapter 4 is a multiphase
study of outflows that looks at the neutral atomic, ionized, and molecular phases of
the ISM.
In Chapter 2, we study molecular outflows and inflows in a combined sample
of local Burst Alert Telescope-detected AGN (BAT AGN)+ULIRGs+QSOs. We
measure the wind detection rate in the BAT AGN and compare it with the rate
of detection in the higher luminosity ULIRGs+QSOs. We also explore what rela-
tionships exist between the properties of the outflows and the properties of their
host galaxies. We then perform statistical tests to measure the strength of the
correlations.
In Chapter 3, we analyze Spitzer -IRS data of 15 U/LIRGs and constrain the
OH-to-H abundance ratio, XOH, by taking advantage of the optically thin OH 35
?m transition in the MIR. We compare our value with those in the literature and
evaluate the accuracy of radiative transfer models to predict absorption line profile
properties.
Chapter 4 explores the multiphase nature of outflows. We analyze Herschel?PACS
data of seven nearby galaxies with well-known winds. We mainly study the neutral
atomic and ionized gas phases, but also include the molecular phase. After delin-
eating the winds from their host galaxy disks, we estimate the spatial (e.g. location,
extent, morphology) and kinematic (e.g. radial velocities, 1-? line widths) proper-
20
ties of the neutral atomic and ionized gas phases in the outflows. We compare our
results with data in other wavelengths. We also present the results of the analysis
of the molecular gas traced by OH 119 ?m.
In Chapter 5, we summarize the main results of each chapter in this thesis and
discuss future work.
21
Chapter 2: Search for Molecular Outflows in Local Volume AGN
with Herschel -PACS
2.1 Introduction
Massive, galactic-scale outflows driven by star formation and/or active galactic
nuclei (AGNs) may be the dominant form of feedback in galaxies (Fabian, 2012;
Veilleux et al., 2005). These outflows (or winds) likely affect the evolution of galaxies
by regulating star formation and black hole (BH) accretion activity. These winds
may shut off the gas feeding process and stop the growth of both the BH and
the spheroidal component (Di Matteo et al., 2005), thereby explaining the tight
?bulge?BH mass relation? (e.g. Fabian, 2012; Kormendy & Ho, 2013; Kormendy
& Richstone, 1995; Marconi & Hunt, 2003). They may also quench star formation
altogether and help explain the presence of ?red-and-dead,? gas-poor ellipticals, and
the bimodal color distribution observed in large galaxy surveys (e.g. Baldry et al.,
2004; Strateva et al., 2001). Winds may also be the primary mechanism by which
metals are transferred from galaxies to their surrounding halos and, to a lesser
extent, to the intergalactic medium.
22
Table 2.1. Herschel Observations of BAT AGN
Name OBSID texp [sec] Program
(1) (2) (3) (4)
CenA 1342225989 976 OT1 shaileyd 1
Circinus 1342225147 958 OT1 shaileyd 1
ESO 005?G004 1342245457 746 OT2 sveilleu 6
ESO 137?34 1342252089 1452 OT2 sveilleu 6
IC 5063 1342241848 2897 OT2 sveilleu 6
IRAS 04410+2807 1342249997 2897 OT2 sveilleu 6
IRAS 19348?0619 1342241495 1452 OT2 sveilleu 6
MCG?05.23.16 1342245454 14251 OT2 sveilleu 6
MCG?06.30.15 1342247813 14251 OT2 sveilleu 6
Mrk18 1342253541 4309 OT2 sveilleu 6
NGC 1052 1342247734 14251 OT2 sveilleu 6
NGC 1068 1342191154 3944 SHINING target
NGC 1125 1342247722 4309 OT2 sveilleu 6
NGC 1365 1342247546 746 OT2 sveilleu 6
NGC 1566 1342244440 746 OT2 sveilleu 6
NGC 2110 1342250314 2158 OT2 sveilleu 6
NGC 2655 1342246552 1452 OT2 sveilleu 6
NGC 2992 1342246246 746 OT2 sveilleu 6
NGC 3079 1342221391 8045 DDT esturm 4
NGC 3081 1342245955 2897 OT2 sveilleu 6
NGC 3227 1342197796 2923 GT1 lspinogl 4
NGC 3281 1342248310 746 OT2 sveilleu 6
NGC 3516 1342245980 3128 GT1 lspinogl 6
NGC 3718 1342253721 7129 OT2 sveilleu 6
NGC 3783 1342247816 2897 OT2 sveilleu 6
NGC 4051 1342247512 1452 OT2 sveilleu 6
NGC 4102 1342247002 746 OT2 sveilleu 6
NGC 4138 1342256950 1452 OT2 sveilleu 6
NGC 4151 1342247511 746 OT2 sveilleu 6
NGC 4258 1342257242 746 OT2 sveilleu 6
NGC 4388 1342197911 3453 GT1 lspinogl 4
NGC 4395 1342247533 1452 OT2 sveilleu 6
NGC 4579 1342248535 746 OT2 sveilleu 6
NGC 4593 1342248372 1452 OT2 sveilleu 6
NGC 4939 1342248509 1452 OT2 sveilleu 6
NGC 4941 1342248508 2862 OT2 sveilleu 6
NGC 4945 1342247792 958 OT1 shaileyd 1
NGC 5033 1342247011 746 OT2 sveilleu 6
NGC 5273 1342246800 14251 OT2 sveilleu 6
NGC 5290 1342247535 2158 OT2 sveilleu 6
NGC 5506 1342247811 746 OT2 sveilleu 6
NGC 5728 1342249309 2508 GT1 lspinogl 6
NGC 5899 1342247007 746 OT2 sveilleu 6
23
Table 2.1 (cont?d)
Name OBSID texp [sec] Program
(1) (2) (3) (4)
NGC 6221 1342252088 746 OT2 sveilleu 6
NGC 6300 1342253357 746 OT2 sveilleu 6
NGC 6814 1342241849 746 OT2 sveilleu 6
NGC 7172 1342218490 2933 GT1 lspinogl 4
NGC 7213 1342245961 1452 OT2 sveilleu 6
NGC 7314 1342245225 746 OT2 sveilleu 6
NGC 7465 1342245963 1452 OT2 sveilleu 6
NGC 7479 1342258846 746 OT2 sveilleu 6
NGC 7582 1342257273 746 OT2 sveilleu 6
Note. ? Column 1: Galaxy name. Column 2: Obser-
vation ID. Column 3: Exposure time. Column 4: Pro-
gram.
10 10 8
BAT AGN
ULIRG/PG QSO
8 8 6
6 6
4
4 4
2
2 2
0 0 0
0.0 0.005 0.01 0.0 0.05 0.1 0.15 0.2 8 9 10 11 12 13
z z log(M?/M)
12 24 12
10 20 10
8 16 8
6 12 6
4 8 4
2 4 2
?0 01 0 1 2 3 0
? 0 20 40 60 80 100
8 9 10 11 12 13
log(SFR/[M yr 1]) ?AGN log(LAGN/L)
Figure 2.1 Histograms showing the distributions of the BAT AGN and ULIRG/PG
QSO properties: redshifts, stellar masses, star formation rates (SFR), AGN frac-
tions, and AGN luminosities.
24
Number of Sources Number of Sources
Until recently, searches for galactic-scale outflows have focused on the brightest
sources in bands often affected by obscuration and/or contamination from the host
galaxy light (e.g. Lehnert & Heckman, 1996). These searches have generally been
directed at either the ionized phase (e.g. Crenshaw et al., 1999; Dunn et al., 2008;
Rubin et al., 2010) or the neutral phase (e.g. Heckman et al., 2000; Krug et al., 2010;
Martin, 2005; Rupke et al., 2005a,b,c; Rupke & Veilleux, 2011, 2013; Schwartz &
Martin, 2004) of the ISM. If winds are to inhibit star formation in the host galaxy,
then the mass outflow must affect the phase of the ISM from which stars form (i.e.
the cold molecular gas). Our knowledge of molecular outflows is quickly improving.
Recent studies of galactic-scale winds have effectively demonstrated that far infrared
(FIR) spectroscopy of the hydroxyl molecule (OH) with Herschel -PACS is well suited
to identify molecular outflows in the nearby universe (Spoon et al., 2013; Sturm
et al., 2011; Veilleux et al., 2013, hereafter, S11, V13, and S13 respectively). S11
reported preliminary evidence of a correlation between AGN luminosity (LAGN) and
OH terminal velocity in six ultra-luminous infrared galaxies (ULIRGs; i.e. ULIRGs
with higher terminal velocities hosted AGNs with higher luminosities). V13 and
S13 later confirmed this correlation via the analyses of larger samples (43 and 24
ULIRGs, respectively). In particular, V13 reported a nonlinear relationship between
log(LAGN/L) and outflow velocity, but noted that better statistics were required
at lower AGN luminosities in order to confirm this nonlinearity. These Herschel -
based studies, supplemented with millimeter-wave, interferometric studies, have also
shown that the molecular gas often dominates the mass and energy budget of these
outflows (e.g. Alatalo et al., 2011; Cicone et al., 2014; Feruglio et al., 2010; Fischer
25
et al., 2010; Gonza?lez-Alfonso et al., 2014b; Morganti et al., 2013; Sturm et al., 2011;
Veilleux et al., 2013).
There is thus a clear need to extend this type of study to lower AGN luminosi-
ties and star formation rates (SFRs). For this, we examine Herschel observations of
a complete sample of local Swift-BAT selected AGNs. Since stellar processes con-
tribute negligibly to the 14-195 keV emission, the Burst Alert Telescope detected
active galactic nucleus (BAT AGN) survey is not sensitive to star formation activity
within the host galaxy. Additionally, this survey is unbiased to column densities of
N . 1024H cm?2. The characteristics and host galaxy properties of these ultra-hard
X-ray detected BAT AGNs have been studied extensively across most wavelengths
(e.g. Koss et al., 2013, 2011; Matsuta et al., 2012; Mele?ndez et al., 2014; Mushotzky
et al., 2014; Shimizu et al., 2016, 2015; Vasudevan & Fabian, 2009; Vasudevan et al.,
2010; Winter et al., 2012). By combining the results of our analysis on 52 BAT AGNs
with those of V13 on 43 ULIRGs, we extend the range of AGN luminosities, SFRs,
and stellar masses sampled in this study by 1-2 orders of magnitude, from which
we can draw stronger statistical conclusions on the driving mechanisms of these
outflows. We also examine mid-infrared (MIR; 5?37 ?m) spectra from the Infrared
Spectrograph (IRS) on board the Spitzer Space Telescope of the combined BAT
AGN + ULIRG + PG QSO sample in an attempt to constrain the distributions
of the dust, as measured by the strength of the 9.7 ?m silicate feature, and OH
gas within these systems. We note that although robust, statistical conclusions can
only be drawn from a well-defined sample, valuable insight into molecular winds
may still be obtained by combining these two samples which are distinct in their
26
0.5 1.5 0.2
CenA ESO 005-G004
1.0 Circinus 0.1
0.0
0.5 0.0
?0.5 0.0 ?0.1
?1.0 ?0.5 ?0.2
?1.0 ?0.3
?1.5
?1.5 ?0.4
?2?.0 ? ?2.0 ?0.53000 2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.15 0.2 0.20
IC 5063 0.16
0.10 ESO 137-34 IRAS 04410+28070.1 0.12
0.05 0.08
0.0 0.04
0.00 0.00
?0.1 ?0.04
?0.05 ?0.08
?0.10 ?0.2 ?0.12
?0.16
?0.?15 ? ? ?0.33000 2000 1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.150 0.10
0.10
0.125
IRAS 19348-0619 MCG-05.23.160.100 MCG-06.30.15
0.05
0.075 0.05
0.050
0.025 0.00 0.00
0.000
?0.025
? ?0.05 ?0.050.050
?0.075
? ? ? ?0.?10 ? ? ?0.103000 2000 1000 0 1000 2000 3000 3000 2000 1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.3 7
0.04
0.2 Mrk18 NGC 1052
6
5 NGC 1068
0.1 0.02
4
0.0 0.00 3
? 20.1 ?0.02
1
?0.2
?0.04 0
?0?.3 ?13000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
Figure 2.2 Fits to the central spaxel, continuum subtracted OH 119 ?m profiles of
the 52 objects in our sample; see Section 2.5.1.3. In each figure, the solid black
line represents the data and the magenta line is the best fit to the data. The blue
and red dashed lines represent the Gaussian components of the fits. The origin of
the velocity scale corresponds to OH 119.233 ?m at the systemic velocity. The two
vertical dashed and dotted lines mark the positions of 16OH and 18OH, respectively.
selection methods and in their properties.
The samples used in this analysis are described in Section 2. The observations
27
Jy Jy Jy Jy
Jy Jy JyJy
Jy Jy
Jy Jy
0.20 0.6 0.4
0.15 NGC 1125
0.4 NGC 1365 0.3
0.10 NGC 1566
0.2
0.05 0.2
0.00 0.1
?0.05 0.0
0.0
?0.10 ?0.2
? ?0.10.15
?0.?20 ? ? ?0.4 ?0.23000 2000 1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.25 0.10
0.1 NGC 29920.20 NGC 2655
NGC 2110
0.15 0.05 0.0
0.10
?0.1
0.05 0.00
0.00 ?0.2
?0.05 ?0.05
?0.3
?0.10
?0.?15 ? ? ?0.?10 ? ? ?0.43000 2000 1000 0 1000 2000 3000 3000 2000 1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
2 0.20 1.0
NGC 3079
1
0.15 0.8
0 NGC 3081 NGC 3227
0.6
?1 0.10
?2 0.4
0.05
?3 0.2
?4 0.00
0.0
?5 ?0.05
? ?0.26
??73000 ? ?0.102000 ?1000 0 1000 2000 3000 ?3000 ? ?0.42000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.3 0.15
0.2 NGC 3281
0.2 NGC 3516 0.10
NGC 3718
0.0
0.05
?0.2 0.1
0.00
?0.4 0.0
?0.05
?0.6 ?0.1 ?0.10
?0.8
? ? ?0.2 ?0.153000 2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
Figure 2.2 (Continued)
and data reduction techniques are outlined in Section 3. The results of the analysis
are presented in Section 4, while the implications of these results are reported in
Section 5. The conclusions are summarized in Section 6.
28
Jy Jy Jy Jy
Jy
Jy Jy Jy
Jy Jy Jy Jy
0.10 0.3
0.5
NGC 4102
NGC 3783 0.2 NGC 4051
0.05
0.0
0.1
0.00
0.0 ?0.5
?0.05 ?0.1
?1.0
?0.?10 ?0.23000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.10 0.25 0.15
NGC 4138 0.20 0.10 NGC 4258NGC 4151
0.05 0.15
0.05
0.10
0.00 0.05 0.00
0.00 ?0.05
?0.05 ?0.05
? ?0.100.10
?0.?10 ?0.15 ?0.153000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.3 0.20 0.20
NGC 4388 0.15
0.15
0.2 NGC 4395 NGC 4579
0.10 0.10
0.1
0.05 0.05
0.0 0.00 0.00
? ?0.05 ?0.050.1
?0.10 ?0.10
?0.2 ?0.15 ?0.15
?0?.3 ?0.20 ?0.203000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.25 0.15 0.15
0.20
0.10 NGC 4939 NGC 4941NGC 4593 0.10
0.15
0.05 0.05
0.10
0.05 0.00 0.00
0.00 ?0.05 ?0.05
?0.05
? ?0.10 ?0.100.10
?0.?153000 ?2000 ? ?0.15 ?0.151000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
Figure 2.2 (Continued)
2.2 BAT AGN Sample Selection
The BAT AGN in our sample were selected using three criteria: (1) all targets
are from the very hard X-ray selected (14-195 keV) 58-month Swift-BAT Survey
(Baumgartner et al., 2011) of local AGN. Since the 14-195 keV flux is solely produced
29
Jy Jy Jy Jy
Jy Jy Jy Jy
Jy Jy Jy Jy
50 0.25 0.10
NGC 4945
0.20 NGC 5273
0 NGC 5033
0.15 0.05
?50 0.10
0.05 0.00
?100 0.00
? ?0.05 ?0.05150
?0.10
?2?00 ? ?0.15 ?0.103000 2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.10 0.6 0.4
NGC 5290 0.4 NGC 5506
0.3 NGC 5728
0.05 0.2
0.2
0.1
0.00 0.0 0.0
? ?0.10.2
?0.05 ?0.2
?0.4 ?0.3
?0.?103000 ?2000 ? ?0.6 ?0.41000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.15 0.25
0.4
NGC 5899 0.20 NGC 6300
0.10 NGC 6221
0.15 0.2
0.05 0.10 0.0
0.05
0.00 ?0.2
0.00 ?0.4
?0.05 ?0.05
? ?0.60.10
?0.10
?0.15 ?0.8
?0.?15 ? ? ?0.?20 ?1.03000 2000 1000 0 1000 2000 3000 3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.20 0.4 0.20
0.15 NGC 7172NGC 6814 0.150.2 NGC 7213
0.10 0.10
0.0
0.05 0.05
0.00 ?0.2 0.00
?0.05 ? ?0.050.4
?0.10 ?0.10
? ?0.60.15 ?0.15
?0.?20 ?0.8 ?0.203000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
Figure 2.2 (Continued)
by the AGN and unaffected by the host galaxy light or obscuration along the line-
of-sight (N . 1024H cm?2, this criterion removes any ambiguity as to the power
of the AGN. Thus, the Swift-BAT survey is arguably superior to soft X-ray, UV,
optical, IR, or radio surveys for understanding the role of AGN-driven winds in
galaxies. (2) All targets have a total integrated flux at 120 ?m of Stot120 & 1 Jy so
30
Jy Jy Jy Jy
Jy Jy Jy Jy
Jy Jy Jy Jy
0.2 0.20 0.4
NGC 7314 NGC 74790.15 NGC 7465 0.2
0.1
0.10
0.0
0.0 0.05 ?0.2
0.00
?0.1 ? ?0.40.05
?0.6
? ?0.100.2
?0.15 ?0.8
?0?.33000 ? ? ?0.?20 ? ? ?1.02000 1000 0 1000 2000 3000 3000 2000 1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.5 NGC 7582
0.0
?0.5
?1.0
?1.5
?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1]
Figure 2.2 (Continued)
that high S/N in the continuum can be reached in a reasonable amount of time
with PACS. (3) Finally, all targets are located within 50 Mpc. The low redshifts in
this sample provide the best possible scale (? 0.02-0.2 kpc arcsec?1) for spatially
separating star-formation emission from the nuclear component. This distance is
also large enough to properly sample the AGN luminosity function up to quasar-like
values (log LBAT ? 43.5; log LBOL ? 45), without favoring IR-bright systems due
to criterion #2.
We find that 52 targets meet these three requirements. Of these targets, 42
objects are from the cycle 2 open-time program OT2 sveilleu 6 (PI: S. Veilleux), 3
objects are from the guaranteed time program GT1 lspinogl 4 (PI: L. Spinoglio),
2 objects are from the guaranteed time program GT1 lspinogl 6 (PI: L. Spinoglio),
3 objects are from the cycle 1 open-time program OT1 shaileyd 1 (PI: S. Hailey-
Dunsheath), 1 object is from the Director?s Discretionary Time DDT esturm 4 (PI:
31
Jy
Jy
Jy
Jy
E. Sturm), and 1 object is from the guaranteed time key program Survey with
Herschel of the ISM in Nearby Infrared Galaxies (SHINING; PI: E. Sturm) (see
Table 2.1).
32
Table 2.2. Galaxy Properties
(Name) z Distance ?AGN log LAGN log M? log SFR fcen/ftot f30?m/f15?m Type
(Mpc) (%) (L) (M ) (M yr
?1
  )
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
CenA 0.001901a 3.7 93.2 9.78 ? ? ? ? ? ? 0.14 2.44 Type 2
Circinus 0.001573a 4.2 86.6 9.19 ? ? ? ? ? ? 0.22 3.32 Type 1
ESO 005?G004 0.006379b 22.4 89.4 9.73 ? ? ? 0.72 0.15 2.95 Type 2
ESO 137?34 0.009144 33 ? ? ? 9.99 ? ? ? 0.5 0.2 ? ? ? Type 2
IC 5063 0.011198c 49 93.3 10.75 ? ? ? 0.61 0.38 2.43 Type 2
IRAS 04410+2807 0.011268 48.7 ? ? ? 10.61 ? ? ? 0.43 0.37 ? ? ? Type 2
IRAS 19348?0619 0.010254 44.3 ? ? ? 10.17 ? ? ? 0.66 0.29 ? ? ? Type 1
MCG?05.23.16 0.008486 36.6 96.9 10.94 9.56 ?0.53 0.59 1.97 Type 2
MCG?06.30.15 0.007749 33.4 99.9 10.36 ? ? ? ?0.4 0.57 1.60 Type 1
Mrk18 0.011345d 47.9 79.0 9.96 9.57 0.41 0.48 4.40 Type 2
NGC 1052 0.005037 19.5 95.7 9.56 10.35 ? ? ? 0.46 2.12 Type 2
NGC 1068 0.003931a 12.7 100 9.26 ? ? ? ? ? ? 0.14 1.20 Type 2
NGC 1125 0.010931 47.2 60.8 10.1 ? ? ? 0.39 0.42 7.31 Type 2
NGC 1365 0.005349a 17.9 75.3 9.82 ? ? ? 1.51 0.18 4.96 Type 1
NGC 1566 0.005017 12.2 ? ? ? 9.01 ? ? ? ? ? ? 0.21 ? ? ? Type 1
NGC 2110 0.007789 35.6 94.0 11.11 10.63 0.49 0.43 2.34 Type 2
NGC 2655 0.00467 24.4 91.8 9.41 ? ? ? ?0.1 0.25 2.63 Type 2
NGC 2992 0.00771 31.6 92.1 9.94 10.31 0.75 0.38 2.59 Type 2
NGC 3079 0.003797d 19.1 68.7 9.59 9.98 1.22 0.29 5.99 Type 2
NGC 3081 0.007976 26.5 93.8 10.27 10.31 0.12 0.35 2.37 Type 2
NGC 3227 0.004001a 18.7 88.6 10.09 9.98 0.55 0.33 3.05 Type 1
NGC 3281 0.010674 46.1 92.7 10.77 10.24 0.84 0.46 2.51 Type 2
NGC 3516 0.008889e 52.5 95.5 11.02 10.46 0.19 0.44 2.15 Type 1
NGC 3718 0.003312 17 ? ? ? 9.05 9.98 ?0.3 0.23 ? ? ? LINER
NGC 3783 0.009791f 47.8 95.2 11.12 ? ? ? 0.68 0.11 2.19 Type 1
33
Table 2.2 (cont?d)
(Name) z Distance ?AGN log LAGN log M? log SFR fcen/ftot f30?m/f15?m Type
(Mpc) (%) (L) (M) (M yr
?1)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
NGC 4051 0.002490a 14.3 95.2 9.41 9.44 0.57 0.34 2.19 Type 1
NGC 4102 0.002823 20.4 61.7 9.57 9.68 1.02 0.51 7.15 LINER
NGC 4138 0.002962 20.7 99.3 9.62 9.61 0.04 0.05 1.67 Type 1
NGC 4151 0.003502a 9.9 100 10.23 ? ? ? ?0.38 0.24 1.48 Type 1
NGC 4258 0.001494 7.5 98.4 8.59 ? ? ? ? ? ? 0.07 1.79 Type 2
NGC 4388 0.008467a 21 84.3 10.6 10.53 0.58 0.14 3.64 Type 2
NGC 4395 0.001064 4.5 72.7 8.21 8.28 ? ? ? 0.1 5.34 Type 1
NGC 4579 0.00506 19.6 ? ? ? 9.11 ? ? ? ? ? ? 0.21 ? ? ? Type 2
NGC 4593 0.008455a 30.8 76.5 10.43 10.75 ? ? ? 0.4 4.78 Type 1
NGC 4939 0.010374 44.8 90.2 10.21 ? ? ? 0.82 0.06 2.84 Type 2
NGC 4941 0.003707d 18.7 89.5 9.36 ? ? ? ?0.19 0.22 2.93 Type 2
NGC 4945 0.001836g 4.1 10.4 9.18 ? ? ? ? ? ? 0.41 18.96 Type 2
NGC 5033 0.003211a 19.6 ? ? ? 8.84 ? ? ? 0.95 0.1 ? ? ? Type 1
NGC 5273 0.003619 16 77.4 9.06 9.64 ? ? ? 0.55 4.64 Type 1
NGC 5290 0.008583 35 ? ? ? 9.88 10.23 0.62 0.12 ? ? ? Type 2
NGC 5506 0.006228a 23.8 93.3 10.64 10.02 0.25 0.52 2.43 Type 1
NGC 5728 0.009475a 30.6 70.4 10.43 10.78 0.78 0.41 5.71 Type 2
NGC 5899 0.008546 38.1 91.0 9.97 10.28 0.95 0.13 2.73 Type 2
NGC 6221 0.004999 12.2 65.4 8.99 ? ? ? 0.76 0.24 6.53 Type 2
NGC 6300 0.003699 14.4 87.0 9.82 ? ? ? 0.63 0.31 3.27 Type 2
NGC 6814 0.005214 22.8 98.0 10.11 10.3 0.6 0.05 1.83 Type 1
NGC 7172 0.009180e 33.9 92.4 10.8 ? ? ? 0.83 0.24 2.54 Type 2
NGC 7213 0.005983d 22 100 9.82 ? ? ? 0.25 0.12 1.46 Type 1
NGC 7314 0.004839h 18.6 89.0 9.77 10.06 ? ? ? 0.13 3.00 Type 1
NGC 7465 0.006666i 27.2 ? ? ? 9.54 ? ? ? 0.21 0.35 ? ? ? Type 2
34
Table 2.2 (cont?d)
(Name) z Distance ?AGN log LAGN log M? log SFR fcen/ftot f30?m/f15?m Type
(Mpc) (%) (L ?1) (M) (M yr )
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
NGC 7479 0.007942 33.9 83.7 9.88 ? ? ? 1.14 0.37 3.73 Type 2
NGC 7582 0.005541j 20.9 73.8 10.06 ? ? ? 1.2 0.49 5.18 Type 2
aRedshift is calculated using the [OI] 63.18 ?m , [OI] 145.53 ?m, [CII] 157.74 ?m, and [NII] 121.90 ?m emission
lines.
bRedshift is calculated using the [OI] 145.53 ?m emission line.
cRedshift is calculated using the [OI] 63.18 ?m and [CII] 157.74 ?m emission lines.
dRedshift is calculated using the [CII] 157.74 ?m emission line.
eRedshift is calculated using the [CII] 157.74 ?m and [NII] 121.90 ?m emission lines.
fRedshift is calculated using the [OI] 63.18 ?m and [OI] 145.53 ?m emission lines.
gRedshift is calculated using the [OI] 63.18 ?m, [OI] 145.53 ?m, and [CII] 157.74 ?m emission lines.
hRedshift is calculated using the [OI] 145.53 ?m and [CII] 157.74 ?m emission lines.
iRedshift is calculated using the [OI] 63.18 ?m, [CII] 157.74 ?m, and [NII] 121.90 ?m emission lines.
jRedshift is calculated using the [OI] 14.53 ?m , [CII] 157.74 ?m, and [NII] 121.90 ?memission lines.
Note. ? Column 1: Galaxy name. Column 2: Redshift value (When available, redshift is calculated from emis-
sion lines. Otherwise, redshifts are from NED). Column 3: When available mean, redshift-independent distance is
obtained from NED. Otherwise, the luminosity distance is calculated via Wright (2006). Column 4: ?AGN, fractional
contribution of the AGN to the bolometric luminosity; see Section 2.4. Column 5: AGN luminosity. Column 6: stellar
masses are adopted from Koss et al. (2011). Column 7: star formation rate; see Section 2.4. Column 8: Continuum
flux density ratio at 119 ?m of the central spaxel to all 25 spaxels. For reference, the average continuum ratio for a
point source calculated from the five PG QSOs in our ULIRG sample is fcen/ftot = 0.56. Column 9: 30 ?m to 15 ?m
continuum flux density ratio. Column 10: Spectral type.
35
2.3 ULIRG and PG QSO Sample
We include in our analysis the Herschel/PACS spectra of the ULIRGs and
PG QSOs from V13. Of the 43 objects (38 ULIRGs + 5 QSOs) from that sample,
23 targets are from the key program SHINING (PI: E. Sturm), 15 targets are from
the cycle 1 open-time program OT1 sveilleu 1 (PI: S. Veilleux), and 5 are from the
cycle 2 open-time program OT2 sveilleu 4 (PI: S. Veilleux).
This sample spans a broad range of merger stages. 20 objects are cool (f25/f60 ?
0.2 ) pre-merger ULIRGs, 18 objects are warm (f25/f60 > 0.2) quasar-dominated,
late-stage, fully coalesced ULIRGs, and 5 objects are ?classic? IR-faint QSOs. These
QSOs are in a late merger phase in which the quasar has finally shed its natal ?co-
coon? of dust and gas and feedback effects may be receding (Veilleux et al., 2009,
V13)
2.4 Properties of the Sample Galaxies
The properties of our BAT AGN sample are listed in Table 2.2. The notes
to Table 2.2 briefly explain the meaning of each of these quantities. Some quan-
tities, however, require further clarification. We apply the bolometric correction
from Winter et al. (2012) to our BAT AGN luminosities, which are derived from
the fluxes from the Swift BAT 70-month survey. Since the Swift-BAT bandpass is
at high enough energies (14-195 keV) to be unaffected by all but the highest levels
of obscuration, the luminosities in this bandpass should be the direct unobscured
36
signature from the AGN. Thus, it is assumed to be a good proxy for the bolometric
luminosity of the AGN. The correction is a scale factor derived from the correlation
between the bolometric luminosity, which is determined from simultaneous broad-
band fitting of the spectral energy distribution of 33 sources in the optical, UV,
and X-ray (Vasudevan & Fabian, 2007, 2009), and the Swift BAT band 14-195 keV
luminosities of those sources. The ordinary least-squares line through these data of
Winter et al. (2012) yields the following correction:
LAGN = 10.5? L14?195keV, (2.1)
where LAGN is the bolometric luminosity of the AGN and L14?195keV is the Swift
BAT luminosity in the 14-195 keV band.
For the ULIRGs, we adopt the starburst and AGN luminosities from V13,
which were calculated as follows: the bolometric luminosities were estimated to be
LBOL = 1.15 LIR, where LIR is the infrared luminosity over 8 ? 1000 ?m (Sanders
& Mirabel, 1996), and LBOL = 7L(5100 A?)+LIR for the PG QSOs (Netzer et al.,
2007). Here, L(5100 A?) corresponds to ?L? at 5100 A?.
The starburst and AGN luminosities were next calculated from
LBOL = LAGN + LSB (2.2)
= ?AGN LBOL + LSB, (2.3)
where ?AGN is the fractional contribution of the AGN to the bolometric luminosity,
37
hereafter called the ?AGN fraction? for short. For the BAT AGN sample, we derive
the AGN fractions from the rest frame f30?m/f15?m continuum flux density ratio,
which was found by Veilleux et al. (2009) to be more tightly correlated with the
PAH-free, silicate-free MIR/FIR ratio and the AGN contribution to the bolometric
luminosity than any other Spitzer-derived continuum ratio. The fraction of the
15 ?m flux produced by the AGN is defined as:
AGN%(f15) (f /f? 30 15
)agn
100 (f30/f15)agn + (f30/f15)sb
(2.4)
(f30/f15)obs ? (f30/f15)= sb? ,(f30/f15)agn (f30/f15)sb
where (f30/f15)obs is the observed flux density ratio. (f30/f15)agn and (f30/f15)sb are
the flux density ratios due to the AGN and starburst, respectively. We adopt from
Table 9 of Veilleux et al. (2009) the zero-point values of log (f30/f15)agn = 0.2 and
log (f30/f15)sb = 1.35.
38
Table 2.3. Properties of the OH 119 ?m Profiles
(Name) v50 (abs) v84 (abs) Fluxabs EQWabs v50 (emi) v84 (emi) Fluxemi EQWemi EQWTotal
(km s?1) (km s?1) (Jy km s?1) (km s?1) (km s?1) (km s?1) (Jy km s?1) (km s?1) (km s?1)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
CenA 63 195 521.1 27 ?201: ?327: ?107.9 ?6 21
Circinus 81 201 494.3 8 ?225 ?327: ?313.9: ?5: 3
ESO 005?G004 9 ?141 167.1 70 622 754 ?32 ?14 55
ESO 137?34 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?9.4 ?14 ?14
IC 5063 ?33:: ?309:: 53.7: 16 ? ? ? ? ? ? ? ? ? ? ? ? 16
IRAS 04410+2807 ? ? ? ? ? ? ? ? ? ? ? ? 21:: 135:: ?36.8:: ?37:: ?37::
IRAS 19348?0619 ? ? ? ? ? ? ? ? ? ? ? ? 57 207 ?28.5 ?25: ?25:
MCG?05?23?016 ? ? ? ? ? ? ? ? ? ? ? ? 57:: 207:: ?8.6:: ?21:: ?21::
MCG?06?30?015 ? ? ? ? ? ? ? ? ? ? ? ? 129:: 255:: ?6.9:: ?8:: ?8::
Mrk 18 ? ? ? ? ? ? 12.4: 6: ? ? ? ? ? ? ? ? ? ? ? ? 6
NGC 1052 ? ? ? ? ? ? ? ? ? ? ? ? ?39:: 117:: ?9.9:: ?27:: ?27::
NGC 1068 ? ? ? ? ? ? ? ? ? ? ? ? 15 213 ?2479.1 ?70 ?70
NGC 1125 123: ?3: 24.8: 20 ? ? ? ? ? ? ? ? ? ? ? ? 20
NGC 1365 ? ? ? ? ? ? ? ? ? ? ? ? 177: 267: ?104.8: ?4: ?4:
NGC 1566 ? ? ? ? ? ? ? ? ? ? ? ? 51 177 ?93.6 ?36 ?36
NGC 2110 ? ? ? ? ? ? ? ? ? ? ? ? 75 351 ?103.9 ?38 ?38
NGC 2655 ? ? ? ? ? ? ? ? ? ? ? ? 9:: 135:: ?8:: ?9:: ?9::
NGC 2992 45 ?129 146.9 33 ? ? ? ? ? ? ? ? ? ? ? ? 33
NGC 3079 87 ?93 2580.2 119 ? ? ? ? ? ? ? ? ? ? ? ? 119
NGC 3081 ? ? ? ? ? ? ? ? ? ? ? ? 93 219 ?38.7 ?33 ?33
NGC 3227 ? ? ? ? ? ? ? ? ? ? ? ? ?27 147 ?225.3 ?53 ?53
NGC 3281 273 429 313.9 99 ?321 ?453 ?32 ?10 ?86
NGC 3516 ? ? ? ? ? ? ? ? ? ? ? ? ?99 39 ?52.4 ?78 ?78
NGC 3718 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?9.3: ?15: ?15:
NGC 3783 ? ? ? ? ? ? ? ? ? ? ? ? ?69:: 81:: ?16.8:: ?33:: ?33::
39
Table 2.3 (cont?d)
(Name) v50 (abs) v84 (abs) Fluxabs EQWabs v50 (emi) v84 (emi) Fluxemi EQWemi EQWTotal
(km s?1) (km s?1) (Jy km s?1) (km s?1) (km s?1) (km s?1) (Jy km s?1) (km s?1) (km s?1)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
NGC 4051 ? ? ? ? ? ? ? ? ? ? ? ? ?9: 141: ?102.2: ?57: ?57:
NGC 4102 ?3 ?135 343.4 12: ? ? ? ? ? ? ? ? ? ? ? ? 12:
NGC 4138 ? ? ? ? ? ? ? ? ? ? ? ? ?99:: 33:: ?19.2:: ?81:: ?81::
NGC 4151 ? ? ? ? ? ? ? ? ? ? ? ? 75 333: ?96.7 ?82 ?82
NGC 4258 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?19.2 ?18 ?18
NGC 4388 ? ? ? ? ? ? 32.7 18 ? ? ? ? ? ? ? ? ? ? ? ? 18
NGC 4395 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?14.2:: ?254:: ?254::
NGC 4579 ? ? ? ? ? ? ? ? ? ? ? ? ?15:: 111:: ?40.3:: ?36:: ?36::
NGC 4593 ? ? ? ? ? ? ? ? ? ? ? ? 15 141 ?44.4 ?26: ?26:
NGC 4939 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?11.1 ?45 ?45
NGC 4941 ? ? ? ? ? ? ? ? ? ? ? ? 3:: 129:: ?22.4:: ?40:: ?40::
NGC 4945 87 ?51 66839.9 171 ? ? ? ? ? ? ? ? ? ? ? ? 171
NGC 5033 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?15.1 ?4 ?4
NGC 5273 ? ? ? ? ? ? ? ? ? ? ? ? 87: 231: ?13.6: ?30: ?30:
NGC 5290 9:: ?117:: 14:: 25:: ? ? ? ? ? ? ? ? ? ? ? ? 25::
NGC 5506 45:: ?357:: 234: 59: ? ? ? ? ? ? ? ? ? ? ? ? 59:
NGC 5728 ?15:: ?141:: 96:: 20:: ? ? ? ? ? ? ? ? ? ? ? ? 20::
NGC 5899 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?11: ?15: ?15:
NGC 6221 ? ? ? ? ? ? ? ? ? ? ? ? ?99:: 33:: ?32:: ?4:: ?4::
NGC 6300 ?3 ?201 301.5 70 ? ? ? ? ? ? ? ? ? ? ? ? 70
NGC 6814 ? ? ? ? ? ? ? ? ? ? ? ? 105:: 231:: ?33.8:: ?82:: ?82::
NGC 7172 ?51: ?207:: 252.6 85 ? ? ? ? ? ? ? ? ? ? ? ? 85
NGC 7213 ? ? ? ? ? ? ? ? ? ? ? ? ?3:: 141:: ?28.3:: ?42:: ?42::
NGC 7314 ? ? ? ? ? ? 18 24 ? ? ? ? ? ? ? ? ? ? ? ? 24
NGC 7465 ? ? ? ? ? ? ? ? ? ? ? ? ?69:: 57:: ?25.6:: ?13:: ?13::
40
Table 2.3 (cont?d)
(Name) v50 (abs) v84 (abs) Fluxabs EQWabs v50 (emi) v84 (emi) Fluxemi EQWemi EQWTotal
(km s?1) (km s?1) (Jy km s?1) (km s?1) (km s?1) (km s?1) (Jy km s?1) (km s?1) (km s?1)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
NGC 7479 ?51 ?658:: 379.6 73 ? ? ? ? ? ? ? ? ? ? ? ? 73
NGC 7582 105 3 383.6 13 604: 700:: ?50.1:: ?2 11
Note. ? Column 1: galaxy name. Column 2: v50(abs) is the median velocity of the fitted absorption profile i.e. 50% of the
absorption takes place at velocities above - more positive than - v50. Column 3: v84(abs) is the velocity above which 84% of the
absorption takes place. Column 4: the total integrated flux for the absorption component(s). Column 5: the total equivalent
width for the absoprtion component(s). Column 6: v50 (emi) is the median velocity of the fitted emission profile. Column 7:
v84 (emi) is the velocity below which 84% of the emission takes place. Column 8: the total integrated flux for the emission
component(s). Column 9: the total equivalent width for the emission component(s). Column 10: The total equivalent width
for the sum of the two components for one line of the OH doublet. Fluxes followed by a colon indicate uncertainties between
20% and 50%. Velocities followed by a colon indicate uncertainties between 50? 150 km s?1. Fluxes followed by a double colon
indicate uncertainties > 50%. Velocities followed by a double colon indicate uncertainties > 150 km s?1.
41
The fraction of the bolometric luminosity produced by the AGN is then cal-
culated from:
? AGN(Lbol) L(bol)
agn
?AGN =
100 L(bol)agn + L(bol)sb
[ ][1( ) ] (2.5)=
(L? (15?m)/L(bol)1 + 100 ? 1 )agn
AGN% (f15) L? (15?m)/L(bol) sb
where we adopt the bolometric corrections of
log [L? (15?m)/L(bol)]agn = ?14.33
log [L? (15?m)/L(bol)]sb = ?14.56
from Table 10 of Veilleux et al. (2009). As a check, we compare the AGN fractions
derived via this method with those found in Tommasin et al. (2010). For the few
objects in common (i.e. NGC 1125, NGC 3516, NGC 4388, NGC 7172, and NGC
7582), we find good agreement between the two methods. For the ULIRGs and
QSOs, we adopt the AGN fractions from V13 which are calculated using the same
method applied to the BAT AGN.
BAT AGN stellar masses are adopted from Koss et al. (2011). ULIRG stellar
masses are calculated by adopting H-band absolute magnitudes from V13. We then
assume M?H = ?23.7 (the H-band absolute magnitude of a L? galaxy in a Schechter
function description of the local field galaxy luminosity function (Bell & de Jong,
2001; Cole et al., 2001)) and m? = 1.4? 1011 M (the mass of an early-type galaxy
42
at the knee of the Schechter distribution (Cole et al., 2001; Veilleux et al., 2002)).
BAT AGN SFRs are derived from equation [25] of Calzetti et al. (2010):
L(160) [erg s?1]
SFR(160)(M yr
?1) = ? , (2.6)4.8 1042
where L(160) is ?L? at 160 ?m and the factor 4.8?1042 assumes a Kennicutt (1998)
calibration for SFR. SFRs for ULIRGs and PG QSOs are calculated from equation
[1] of Rupke et al. (2005b):
LSB
SFR = ? , (2.7)5.8 109 L
where LSB = LBOL(1? ?AGN).
Table 2.1 shows the distributions of redshifts, stellar masses, SFRs, AGN
fractions, and AGN luminosities for all 52 BAT AGN in our sample (blue diagonals)
and for all 43 objects (38 ULIRGs + 5 PG QSOs; red diagonals) from V13. The
AGN fractions of the BAT AGN are typically higher than the fractions of the V13
sample of objects.The AGN luminosities and SFRs of the BAT AGN, however, are
typically lower by two orders of magnitude than those of the ULIRGs and QSOs.
43
2.5 Observations, Data Reduction, and Spectral Analysis
2.5.1 OH 119 ?m Feature
2.5.1.1 OH Observations
The OH observations were obtained with the PACS FIR spectrometer (Poglitsch
et al., 2010) on board Herschel (Pilbratt et al., 2010). We focus our efforts on the
ground-state OH 119 ?m 2?3/2 J = 5/2? 3/2 rotational ?-doublet. This feature is
the strongest transition in ULIRGs and is positioned near the peak spectroscopic
sensitivity of PACS. Fischer et al. (2010), S11, and V13 have demonstrated the
efficacy of using the OH 119 ?m feature to determine wind characteristics.
The data for OT2 sveilleu 6 were obtained in a similar fashion as those in
SHINING (S11) and OT1 sveilleu 1 (V13). PACS was used in range scan spec-
troscopy mode in high sampling centered on the redshifted OH 119 ?m + 18OH 120
?m complex with a velocity range of ? ?4000 km s?1 (rest-frame 118-121 ?m) to
provide enough coverage on both sides of the OH complex for reliable continuum
placement. As a result, the PACS spectral resolution is ? 270 km s?1. The to-
tal amount of observation time (including overheads) for OT2 sveilleux 6 was 35.3
hours. The program ID and observing time for each target, including all overheads,
are listed in Table 2.2. A large chopper throw of 3? is used in all cases.
44
300 Abs (BAT AGN)(a) Emi (BAT AGN) (b) (c)
Abs/Emi (BAT AGN)
Abs (ULIRG)
Emi (ULIRG)
200 Abs/Emi (ULIRG)
Emi (PG QSO)
100
0
?100
9 10 11 12 13 0 20 40 60 80 100 8 9 10 11 12 13
log(M?/M) ?AGN [%] log(LAGN/L)
Figure 2.3 Total (absorption + emission) equivalent widths of OH 119 ?m (positive
values indicate absorption and negative values indicate emission) as a function of
the (a) stellar masses, (b) AGN fractions, (c) AGN luminosities. The colors blue
and red refer to BAT AGN and ULIRGs/PG QSOs, respectively. Filled squares,
triangles, and circles represent BAT AGN or ULIRGs in which OH 119 ?m is seen
purely in absorption, purely in emission, or with composite absorption/emission,
respectively. Open triangles represent PG QSOs in which OH 119 ?m is seen purely
in emission. The blue, vertical, dotted lines refer to BAT AGN in which OH 119 ?m
is not detected. Similarly, the red, vertical, dash-dotted lines represent ULIRGs/PG
QSOs with undetected OH 119 ?m. The gray, horizontal line marks the null OH
equivalent width.
2.5.1.2 OH Data Reduction
The reduction of the OH data on the BAT AGN sample was carried out using
the standard PACS reduction and calibration pipeline (ipipe) included in HIPE 6.0.
For the final calibration, fainter sources were normalized to the telescope flux (which
dominates the total signal) and recalibrated using a reference telescope spectrum
obtained from dedicated Neptune observations during the Herschel performance
verification phase. Sturm et al. (2011) demonstrated that this telescope background
technique can reliably recover the continuum from faint sources. In the following,
we use the spectrum of the central 9.4?? ? 9.4?? spatial pixel (spaxel) only, without
the application of a point source flux correction. In a few objects we note that the
45
EW [km s?1OH ]
OH emission is extended. These objects will be the focus of a future paper.
The reduced spectra were next smoothed using a Gaussian kernel of width
0.05 ?m (i.e. about half a resolution element) to reduce the noise in the data before
the spectral analysis. A spline was fit to the continuum emission and subtracted
from the spectra. Subsequently, spectral fitting was carried out on these continuum-
subtracted spectra.
2.5.1.3 Spectral Analysis of the OH Doublet
Line profile fits of the OH 119.233, 119.441 ?m doublet were computed by
using PySpecKit, a spectroscopic analysis and reduction toolkit for optical, infrared,
and radio spectra (Ginsburg & Mirocha, 2011). The toolkit uses the Levenberg-
Marquardt technique to solve the least-squares problem in order to find the best fit
for the observations. Profile fitting of the OH doublet followed a similar procedure as
that outlined in V13, in which the doublet profile was modeled using four Gaussian
components (two components for each line of the doublet), each characterized by
their amplitude (either negative or positive), peak position, and standard deviation
(or, equivalently FWHM). However, many of the OH profiles observed here were fit
with only two Gaussian components (one component for each line of the doublet),
since a fit with four Gaussian components often led to spurious results (i.e. skinny
components with FWHM ? 10 km s?1, values too small to be considered real). The
separation between the two lines of the doublet was set to 0.208 ?m in the rest
frame (? 520 km s?1) and the amplitude and standard deviation were fixed to be
46
the same for each component in the doublet. In cases where OH was not detected,
two Gaussian components, characterized by an amplitude consistent with the 1?
level of the noise and a FWHM (300 km s?1) approximately equal to the resolution
of PACS, were fit and maximum values for the OH flux and equivalent width were
derived.
Four distinct scenarios apply to our data: (1) pure OH absorption, (2) pure
OH emission, (3) P Cygni profiles, and (4) inverse P Cygni profiles. In scenario 1,
there is no evidence of OH emission and each line of the OH doublet is fitted with 1-2
absorption components. In the case of the two-component fit, one component traces
the stronger low-velocity component of the outflow, while the other component
traces the fainter high-velocity component. In the case of the single-component
fit, only the low-velocity component is captured. Scenario 2 is treated similarly.
In this scenario, there is no evidence for any OH absorption and 1-2 Gaussian
components are used to model each line of the OH doublet. In scenario 3, each line
of the doublet is modeled with a single blueshifted absorption and a single redshifted
emission component. In scenario 4, each line of the doublet is modeled with a single
blueshifted emission component and a single redshifted absorption component. In
scenarios 3 and 4, just as in the one-component fits of Scenarios 1 and 2, only the
low-velocity component of the outflow is captured.
These fits were first used to quantify the strength and nature (absorption
versus emission) of the OH feature: (1) the total flux and equivalent width of the
OH 119.441 ?m line, adding up all of the absorption and emission components, (2)
the flux and equivalent width of the absorption component(s) used to fit this line,
47
and (3) the flux and equivalent width of the emission component(s) used to fit this
line.
Following the same method as that outlined in V13, we also characterize the
OH profile by measuring velocities: (1) v50(abs) is the median velocity of the fitted
absorption profile, i.e., 50% of the absorption takes place at velocities above (more
positive than) v50(abs), (2) v84(abs) is the velocity above which 84% of the absorption
takes place, (3) v50(emi) is the median velocity of the fitted emission profile, i.e.,
50% of the emission takes place at velocities below (less positive than) v50(emi),
and (4) v84(emi) is the velocity below which 84% of the emission takes place. We
note that three objects are fitted with inverted P-Cygni profiles, suggesting inflow.
Circinus is an unambiguous fit while Centaurus A and NGC 3281 are marginally fit
with inverted P-Cygni profiles. The velocities from the inverted P-Cygni profile fits
are measured as thus: v50(abs) and v84(abs) are the velocities below which 50% and
84% of the absorption takes place, respectively, and v50(emi) and v84(emi) are the
velocities above which 50% and 84% of the emission takes place, respectively.
We note that the OH emission is extended in some of our BAT AGN sources.
While a full analysis of the 5 ? 5 spaxels spectrum is outside the scope of this paper,
we provide the ratio of the central spaxel 119 ?m continuum flux density to that
of the summed 25 spaxels (fcen/ftot) to quantify the degree to which some of our
BAT AGN are spatially extended (see Table 2.2). For reference, the average value
of fcen/ftot for a point source (derived from the five PG QSOs in our sample) is 0.56.
Note that the continuum is often considerably more extended than the OH 119 ?m
feature.
48
8
Abs (BAT AGN)
7 Emi (BAT AGN)
Abs/Emi (BAT AGN)
Null (BAT AGN)
6 Abs (ULIRG)
Emi (ULIRG)
Abs/Emi (ULIRG)
5 Null (ULIRG)
Emi (PG QSO)
4
3
2
1
0
?1
0 20 40 60 80 100
?AGN [%]
Figure 2.4 The apparent strength of the 9.7 ?m silicate feature relative to the local
mid-infrared continuum as a function of the AGN fractions. Note that S9.7?m is a
logarithmic quantity and can be interpreted as the apparent silicate optical depth.
The strength of silicate absorption increases upward. Sign conventions and meanings
of the symbols are the same as those in Section 2.5.1.1. Crosses represent objects
with a null OH detection. Vertical lines indicate objects with a null 9.7 ?m silicate
feature detection.
2.5.2 The 9.7 ?m Silicate Feature
2.5.2.1 Data Reduction of the 9.7 ?m Silicate Feature
Mid-infrared (MIR; 5-37 ?m) low resolution spectra (R ? 60 ? 127) used
to measure the 9.7 ?m silicate feature for the BAT AGN, ULIRG, and PG QSO
samples were extracted from The Cornell AtlaS of Spitzer/IRS Sources (CASSIS1).
The archival data were obtained with the Infrared Spectrograph (IRS; Houck et al.
(2004)) on board the Spitzer Space Telescope (Werner et al., 2004). There were
two objects however (NGC 7479 and NGC 4945), in which the MIR spectrum was
1http://www.cassis.sirtf.com/atlas
49
S9.7?m
extracted from the Spitzer Heritage Archive (SHA2).
The Spitzer observations were made with the Short-Low (SL, 5.2-14.5 ?m)
and Long-Low (LL, 14.0-38.0 ?m) modules of the IRS. The orders were stitched to
LL order 2, requiring order-to-order scaling adjustments of less than ? 15%.
2.5.2.2 Spectral Analysis of the 9.7 ?m Silicate Feature
We have measured the strength of the 9.7 ?m silicate feature for sources in
our BAT AGN sample and in the ULIRG/QSO sample. The calculation of the mid-
infrared continuum loosely follows the method of Spoon et al. (2007). For a majority
of the sources, the mid-infrared continuum (e.g. the continuum flux density fcont) is
determined from a cubic spline interpolation to continuum pivots at 5.2, 5.6, 14.0,
27.0, and 31.5 ?m. For objects in which the 9.7 ?m silicate feature dominates the
spectrum (i.e. there is very little PAH emission), an additional pivot point is added
at 7.8 ?m. The pivot points are adopted from Spoon et al. (2007), however we have
placed a pivot point at 27 ?m instead of at 25 ?m due to the proximity of the [OIV]
emission line at 25.89 ?m which is common in BAT AGN. Due to the diversity of
our BAT AGN sample, the wavelength range in which the 9.7 ?m silicate feature is
determined differs slightly from source to source (see Table 2.4). Typically, however,
the observed flux density (e.g. fobs) is determined from a cubic spline interpolation
to the data between ? 8?m and ? 14?m. This interpolation skips the H2 line at
9.6 ?m, [S IV] line at 10.51 ?m, PAH feature at 11.25 ?m, H2 line at 12.28 ?m, and
[NeII] at 12.68 ?m.
2http://sha.ipac.caltech.edu/applications/Spitzer/SHA/
50
The apparent strength of the 9.7 ?m silicate feature is then defined as:
( )
fobs(9.7?m)
S9.7?m = ? ln . (2.8)
fcont(9.7?m)
Here, we adopt the sign convention consistent with our OH 119 ?m analysis
(e.g. positive S9.7?m values indicate absorption while negative S9.7?m values indicate
emission). Note that this convention is different from previous studies (e.g. Spoon
et al., 2007).
For sources with a silicate absorption feature, S9.7?m can be interpreted as
the the apparent silicate optical depth. Table 2.4 and Table 2.5 lists pivot points,
integration ranges, and measured equivalent widths and fluxes of the 9.7 ?m silicate
feature for our samples. MIR spectra showing these pivot points and integration
ranges may be found in the Appendix.
2.6 Results
Figure 2.1 shows the fits to the OH 119 ?m profiles for our BAT AGN targets.
The OH 119 ?m parameters for all targets derived from these fits are listed in
Table 3.3. The meaning of each parameter is discussed in Section 2.5.1.3 and in the
notes to Table 3.3. Note that the fluxes and equivalent widths in Table 3.3 need to
be multiplied by a factor of two when considering both lines of the doublet.
In this section we compare the OH 119 ?m and S9.7?m results listed in Table 3.3,
Table 2.4, and Table 2.5, with the galaxy properties listed in Table 2.2.
51
300
200
100
0
Abs (BAT AGN)
Emi (BAT AGN)
Abs/Emi (BAT AGN)
? Abs (ULIRG)100 Emi (ULIRG)
Abs/Emi (ULIRG)
Emi (PG QSO)
?1 0 1 2 3 4 5
S9.7?m
Figure 2.5 Total (absorption + emission) equivalent widths of OH 119 ?m as a
function of the apparent strength of the 9.7 ?m silicate feature relative to the local
mid-infrared continuum. Note that S9.7?m is a logarithmic quantity and can be
interpreted as the apparent silicate optical depth. The strength of this absorption
feature increases to the right. Sign conventions and meanings of the symbols are
the same as those in Section 2.5.1.1. Horizontal, blue, dotted lines represent BAT
AGN in which the 9.7 ?m silicate feature is not detected. Similarly, horizontal, red,
dash-dotted lines represent ULIRGs/PG QSOs with a null 9.7 ?m silicate feature
detection. The black diagonal line shows the ordinary least squares bisector linear
regression: EWOH = 56.81?S9.7?m?26.7 kms?1. The Pearson r null probability for
the linear relationship between S9.7?m and the total OH equivalent width is P [null]=
1.2 ? 10?9 for the BAT AGN sample and P [null] = 0.003 for the ULIRG + QSO
sample. When the samples are combined, we find P [null] = 3.9? 10?9.
2.6.1 The OH 119 ?m Feature
The OH 119 ?m doublet is detected in 42 of the 52 objects in our BAT AGN
sample. Of the 42 detections, 25 are seen purely in emission, 12 are seen purely in ab-
sorption, and 5 are seen with absorption+emission composite profiles. For compari-
son, in the ULIRG + QSO sample (V13), 37 objects showed OH 119 ?m detections,
52
EW ?1OH [km s ]
14 12
Pure Abs Pure Abs
12 Pure Emi Pure Emi
PCyg 10 PCyg
10 Inv PCyg Inv PCyg8
8
6
6
4 4
2 2
?0800 ? 0400 0 400 800 ?800 ?400 0 400 800
8 8
Pure Abs Pure Abs
PCyg PCyg
Inv PCyg Inv PCyg
6 Total 6 Total
4 4
2 2
?0 0400 ?200 0 200 400 ?800 ?400 0 400
v ?1 ?150 [km s ] v84 [km s ]
Figure 2.6 Histograms showing the distributions of the 50% (median; left panels)
and 84% (right panels) velocities derived from the multi-Gaussian fits to the OH
profiles of the BAT AGN. Top panels show pure absorption components (blue),
pure emission components (red dashed), P-cygni emission components (red, left di-
agonals), and inverse P-cygni emission components (red, right diagonals). Bottom
panels show pure absorption components (filled grey), P-Cygni absorption com-
ponents (blue, left diagonals), inverse P-cygni absorption components (blue, right
diagonals), and total absorption components (pure + P-Cygni + inverse P-Cygni).
17 of which were seen purely in absorption, 15 showed absorption+emission com-
posite profiles, and 5 were seen purely in emission. Section 2.5.1.1a, Section 2.5.1.1c
plot the OH equivalent width with the stellar mass and AGN luminosity, respec-
tively. For these properties, we see no discernible correlation in either the individual
samples (i.e. BAT AGN only and ULIRG + PG QSO only) or in the combined
sample. There is however a weak trend between EWOH and AGN fraction (Sec-
tion 2.5.1.1b) when the samples are combined. A more formal statistical analysis
of these parameters indeed indicates a statistically significant correlation between
53
Number of Sources
these quantities when all objects are considered (see Table 2.6).
For the BAT AGN in which OH 119 ?m is seen purely in emission, we find
that the AGN is dominant (i.e. ?AGN > 50%) in these objects. In agreement with
V13, the low luminosities (log(LAGN/L) . 12) of these 20 AGN-dominated objects
with pure OH emission suggest that the AGN fraction is more important than AGN
luminosity in setting the character (i.e. strength of emission relative to absorption)
of the OH feature. Therefore, we see that dominant AGN in our sample can excite
the OH molecule (via radiative pumping or collisional excitation) and produce the
2?3/2 J = 5/2? 3/2 rotational emission line. This trend was noted in V13.
54
Table 2.4. S9.7?m Continuum Parameters and Measured Values of the BAT AGN
Sample
(Name) Pivot Points Integration Range FluxS9.7?m EQWS9.7?m S9.7?m
(?m) (?m) (Jy ?m) (?m)
(1) (2) (3) (4) (5) (6)
CenA 5.2, 5.6, 14.0, 27.0, 31.5 8.3, 11.3 1.06 1.01 0.79
Circinus 5.2, 5.6, 14.0, 31.5 7.8, 14 23.38 2.06 1.24
ESO 005?G004 5.2, 5.6, 14.0, 27.0, 31.5 8.6, 11.19 0.34 1.55 1.32
ESO 137?34 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
IC 5063 5.2, 5.6, 14.0, 27.0, 31.5 8.15, 12.3 0.37 0.55 0.27
IRAS 00410+2807 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
IRAS 19348?0619 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
MGC?05.23.16 5.2, 5.6, 14.0, 27.0, 31.5 8, 12.45 0.32 0.64 0.31
MGC?06.30.15 5.2, 5.6, 14.0, 27.0, 31.5 8.09, 10.4 0.05 0.16 0.09
Mrk 18 5.2, 5.6, 14.0, 27.0, 31.5 9, 10.4 0.01 0.12 0.11
NGC 1052 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 1068 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 1125 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 1365 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 1566 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 2110 5.2, 5.6, 14.0, 27.0, 31.5 7.8, 14 0.17 0.74 ?0.07
NGC 2655 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 2992 5.2, 5.6, 14.0, 27.0, 31.5 8.75, 11.1 0.13 0.48 0.37
NGC 3079 5.2, 5.6, 14.0, 27.0, 31.5 8.8, 11.13 0.66 1.5 1.56
NGC 3081 5.2, 5.6, 14.0, 27.0, 31.5 8, 11.3 0.09 0.44 0.22
NGC 3227 5.2, 5.6, 14.0, 27.0, 31.5 8.75, 11.09 0.15 0.46 0.28
NGC 3281 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 7.8, 14 1.98 2.27 1.36
NGC 3516 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 3718 ? ? ? ? ? ? ? ? ? ? ? ?
NGC 3783 5.2, 5.6, 14.0, 27.0, 31.5 7.9, 11.3 0.17 0.33 0.13
55
Table 2.4 (cont?d)
(Name) Pivot Points Integration Range FluxS EQW S9.7?m S9.7?m 9.7?m
(?m) (?m) (Jy ?m) (?m)
(1) (2) (3) (4) (5) (6)
NGC 4051 5.2, 5.6, 14.0, 21.9, 24.9 8, 10.8 0.07 0.24 ?0.07
NGC 4102 5.2, 5.6, 14.0, 27.0, 31.5 9, 10.95 0.23 0.52 0.41
NGC 4138 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 4151 ? ? ? 7.8, 14 ? ? ? ? ? ? ? ? ?
NGC 4258 5.2, 5.6, 14.0, 27.0, 31.5 7.8, 13.39 0.14 1.15 ?0.16
NGC 4388 5.2, 5.6, 14.0, 27.0, 31.5 8.2, 12.2 0.43 1.35 0.79
NGC 4395 5.2, 5.6, 14.0, 27.0, 31.5 7.8, 11.09 0.004 0.67 0.25
NGC 4579 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 4593 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 4939 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 4941 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 8, 11.13 0.02 0.33 0.15
NGC 4945 7.8, 14.0, 27.0, 31.5 7.8, 11.09 1.95 2.72 3.65
NGC 5033 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 5273 4.6, 14.0, 27.0, 31.5 7.7, 11.05 0.01 0.66 0.38
NGC 5290 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 5506 5.2, 5.6, 14.0, 27.0, 31.5 8.35, 12.15 1.2 1.16 0.77
NGC 5728 5.2, 5.6, 14.0, 27.0, 31.5 8.63, 11.16 0.15 1.17 0.94
NGC 5899 5.2, 5.6, 14.0, 27.0, 31.5 8.1, 11.15 0.08 1.44 1.02
NGC 6221 5.2, 5.6, 14.0, 27.0, 31.5 9.2, 10.7 0.03 0.13 0.11
NGC 6300 5.2, 5.6, 14.0, 27.0, 31.5 8.58, 12.2 0.67 1.71 1.16
NGC 6814 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 7172 5.2, 5.6, 14.0, 27.0, 31.5 8.15, 12.52 0.68 2.37 1.97
NGC 7213 5.2, 5.6, 14.0, 27.0, 31.5 8.76, 13.3 0.36 2.26 ?0.41
NGC 7314 5.2, 5.6, 7.8, 14.0 7.8, 14 0.09 1.38 0.6
NGC 7465 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
56
Table 2.4 (cont?d)
(Name) Pivot Points Integration Range FluxS9.7?m EQWS9.7?m S9.7?m
(?m) (?m) (Jy ?m) (?m)
(1) (2) (3) (4) (5) (6)
NGC 7479 4.6, 7.8, 14.0, 27.0, 31.5 7.8, 12.9 1.98 3.09 2.3
NGC 7582 5.2, 5.6, 14.0, 27.0, 31.5 8.8, 11.05 0.81 0.95 0.78
Note. ? Column 1: galaxy name. Column 2: Pivot points in microns used for the continuum
interpolation. Column 3: Integration range of the 9.7 ?m silicate feature. Column 4: Total
integrated flux of the 9.7 ?m silicate feature. Column 5: total equivalent width for the 9.7 ?m
silicate feature. Column 6: S9.7?m ; see Section 2.5.2.2.
57
2.6.2 The S9.7?m Feature
Figure 2.4 plots the apparent strength of the 9.7 ?m silicate absorption feature
versus the AGN fraction ?AGN. Here, larger, more positive values of S9.7?m indicate
stronger silicate absorption features. We see no discernible trend between the
strength of this silicate feature and AGN fraction. This is surprising since the
AGN fractions should be inversely proportional to the equivalent widths of the PAH
features (Veilleux et al., 2009). Thus, Figure 2.4 should look similar to Figure 1
of Spoon et al. (2007) which plots the 6.2 ?m PAH emission feature and S9.7?m .
We find that our data do not show the bifurcation observed in Spoon et al. (2007)
because only two of our ULIRGs (F08572+3915 and F15250+3608) lie on the upper
branch of the bifurcation, and none of our BAT AGN populate the upper branch.
The two objects in Figure 2.4 which display ?AGN ? 10% and high S9.7?m are NGC
4945 and F15327+2340. Both of these objects show weak PAH emission and are
therefore not displayed in Figure 1 of Spoon et al. (2007).
Figure 2.5 shows a positive correlation between the measured OH equivalent
width and the 9.7 ?m silicate optical depth. The black diagonal line shows the
ordinary least squares bisector linear regression for the entire sample (BAT AGN +
ULIRGs + PG QSOs):
EWOH = 56.81? S ?19.7?m ? 26.7 [km s ]. (2.9)
The Pearson correlation coefficient of the linear relationship for S9.7?m and the
58
total OH equivalent width is ?S9.7?m,EW = 0.7 with a P [null] (the probabil-OH
ity of an uncorrelated system producing datasets that have a Pearson correlation
at least as extreme as the one computed from these datasets) of 3.9 ? 10?9 for
the combined sample. If we restrict this analysis to the BAT AGN only, we find
?S9.7?m,EW (BAT AGN) = 0.89 with a P [null] of 1.2? 10?9. Restricting the analy-OH
sis to ULIRGs and PG QSOs only yields ?S9.7?m,EW (ULIRG/PG QSO) = 0.5 withOH
a P [null] of 0.003.
In Figure 2.5 we also see that objects with OH in emission show either weak
silicate absorption or silicate emission. Objects with OH P-Cygni profiles show
moderate silicate absorption features, while the strongest silicate absorption features
are seen in objects in which OH is observed purely in absorption. We return to this
result in Section 3.5.
2.6.3 OH Kinematics
We quantify the visual trends observed (or not observed) in our kinematic
investigation of the individual samples (i.e. BAT AGN only or ULIRG/QSO only)
and of the combined sample by calculating the correlational significances between the
observed OH velocities (v50 and v84) and the host galaxy properties (stellar masses,
SFRs, specific star formation rates (sSFRs; i.e. the rate at which stars are formed
divided by the stellar mass of the galaxy), and AGN fractions and luminosities.
Since the spatial location of the OH emission is unknown, physical interpretations
of the observed OH velocities in these cases will be ambiguous (e.g., blueshifted
59
velocities may correspond to outflow or inflow if the OH emission region is located
in front or behind the continuum source, respectively). We therefore exclude from
our analysis objects in which OH is detected purely in emission. The results of our
statistical analyses on all objects with either redshifted or blueshifted absorption
profiles are listed in Table 2.7.
Section 2.6 shows the distributions of velocities derived from both the OH
absorption and emission line features (v50(abs), v84(abs), v50(emi), and v84(emi),
as defined in Section 2.5.1.3 and listed in Table 3.3). In V13, we adopted Rupke
et al. (2005a)?s conservative definition of an outflow as having an OH absorption
feature with a median velocity (v50) more negative than ?50 km s?1. Similarly, we
can define an inflow as having an OH absorption feature with a median velocity
(v50) more positive than 50 km s
?1. This cutoff is used to avoid contamination due
to systematic errors and measurement errors in wavelength calibration, line fitting
(see Section 2.5.1.3), and redshift determination. We find that only two objects in
our BAT AGN sample meet this outflow criterion, and just barely: NGC 7172 and
NGC 7479 (both have v ?150 = ?51 km s ). The presence of a significant blue wing
in the absorption profile of OH 119 ?m in NGC 7479 with v84 = ?658 km s?1 adds
considerable support to this interpretation. An extended blue wing with v84 < ?300
km s?1 may also be present in the OH absorption profile of IC 5063, NGC 5506,
and NGC 7172, although these detections are more tentative than in NGC 7479. In
contrast, seven objects have OH absorption features with median velocities larger
than +50 km s?1: Centaurus A, Circinus, NGC 1125, NGC 3079, NGC 3281, NGC
4945, and NGC 7582. The clear detection of an inverted P-Cygni profile in Circinus
60
provides unambiguous evidence for the presence of an inflow in this object. The
inverted P-Cygni profiles in Centaurus A and NGC 3281 are much less secure. We
return to these objects in Section 3.5.
61
Table 2.5. S9.7?m Continuum Parameters and Measured Values of the
ULIRG/PG QSO Sample
(Name) Pivot Points Integration Range FluxS EQW9.7?m S S9.7?m 9.7?m
(?m) (?m) (Jy ?m) (?m)
(1) (2) (3) (4) (5) (6)
07251?0248 5.2, 14.0, 27.0, 31.5 8.45, 12.7 0.21 3.4 2.37
09022?3615 5.2, 5.6, 14.0, 27.0, 31.5 8.3, 12.6 0.43 1.84 1.22
13120?5453 5.2, 5.6, 14.0, 27.0, 31.5 8.7, 12.0 0.5 1.72 1.49
19542+1110 5.2, 5.6, 14.0, 27.0, 31.5 8.74,11.14 0.05 1.12 0.86
F00509+1225 5.2, 5.6, 14.0, 27.0, 31.5 7.8, 14.0 0.31 0.82 ?0.24
F01572+0009 5.2, 5.6, 14.0, 27.0, 31.5 8.1, 14.0 0.06 0.75 ?0.17
F05024?1941 5.2, 5.6, 10.8, 18.5 7.8, 12.6 0.01 1.65 0.22
F05189?2524 5.2, 5.6, 14.0, 27.0, 31.5 8.37, 14.0 0.54 1.01 0.37
F07598+6508 5.2, 5.6, 14.0, 27.0, 31.5 7.2, 14.0 0.11 0.49 ?0.15
F08572+3915 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 7.8, 13.0 2.55 3.39 4.00
F09320+6134 5.2, 5.6, 14.0, 27.0, 31.5 8.65, 12.1 0.33 1.91 1.58
F10565+2448 5.2, 5.6, 14.0, 27.0, 31.5 8.76, 11.17 0.16 1.21 1.01
F11119+3257 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 7.8, 12.6 0.17 1.21 0.62
F12072?0444 5.2, 5.6, 14.0, 27.0, 31.5 8.1, 12.6 0.22 2.36 1.41
F12112+0305 5.2, 5.6, 14.0, 27.0, 31.5 8.6, 12.4 0.13 2.42 1.58
F12243?0036 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 8.15, 12.8 4.73 3.95 4.03
F12265+0219 5.2, 5.6, 14.0, 27.0, 31.5 8.8, 14.0 0.08 0.26 ?0.06
F12540+5708 5.2, 5.6, 14.0, 27.0, 31.5 8.4, 12.6 2.62 1.49 0.62
F13305?1739 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
F13428+5608 5.2, 5.6, 14.0, 27.0, 31.5 8.2, 12.6 0.73 2.77 2.02
F13451+1232 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 7.8, 12.2 0.06 0.61 0.35
F14348?1447 5.2, 5.6, 14.0, 27.0, 31.5 8.3, 12.4 0.15 2.66 1.93
F14378?3651 5.2, 5.6, 14.0, 27.0, 31.5 8.7, 12.1 0.06 1.76 1.55
F14394+5332 5.2, 5.6, 14.0, 27.0, 31.5 8.5, 12.3 0.18 2.26 1.71
F15206+3342 5.2, 5.6, 14.0, 27.0, 31.5 9.0, 11.1 0.02 0.47 0.29
62
Table 2.5 (cont?d)
(Name) Pivot Points Integration Range FluxS EQW9.7?m S S9.7?m 9.7?m
(?m) (?m) (Jy ?m) (?m)
(1) (2) (3) (4) (5) (6)
F15250+3608 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 7.8, 14.0 0.94 3.54 3.42
F15327+2340 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 7.8, 12.6 2.87 3.57 3.49
F15462?0450 5.2, 5.6, 14.0, 27.0, 31.5 8.6, 12.4 0.06 0.78 0.29
F16504+0228 5.2, 5.6, 14.0, 27.0, 31.5 8.65, 12.2 0.72 2.06 1.54
F17208?0014 5.2, 5.6, 14.0, 27.0, 31.5 8.6, 12.4 0.37 2.38 1.96
F19297?0406 5.2, 5.6, 14.0, 27.0, 31.5 8.6, 12.5 0.12 2.48 1.60
F20551?4250 5.2, 5.6, 7.8, 14.0, 27.0, 31.5 7.8, 14.0 1.52 3.43 3.09
F22491?1808 5.2, 5.6, 14.0, 27.0, 31.5 8.65, 12.4 0.07 2.2 1.14
F23128?5919 5.2, 5.6, 14.0, 27.0, 31.5 8.65, 12.2 0.23 1.45 0.88
F23233+2817 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
F23365+3604 5.2, 5.6, 14.0, 27.0, 31.5 8.65, 12.4 0.17 2.49 1.78
F23389+0300 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
PG 1126?041 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
PG 1351+640 5.2, 5.6, 14.0, 27.0, 31.5 8.1, 14.0 0.24 2.31 ?0.58
PG 1440+356 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
PG 1613+658 5.2, 5.6, 14.0, 27.0, 31.5 7.2, 14.0 0.03 0.38 ?0.01
PG 2130+099 5.2, 5.6, 14.0, 27.0, 31.5 7.75, 14.0 0.05 0.29 0.10
Z11598?0114 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Note. ? Column 1: galaxy name. Column 2: Pivot points in microns used for the continuum inter-
polation. Column 3: Integration range of the 9.7 ?m silicate feature. Column 4: Total integrated flux of
the 9.7 ?m silicate feature. Column 5: total equivalent width for the 9.7 ?m silicate feature. Column 6:
S9.7?m ; see Section 2.5.2.2.
63
Table 2.6. Results from Statistical Analyses of Host Galaxy Properties
Parameter Number of Objects ?s P? ? P? r Pr
(1) (2) (3) (4) (5) (6) (7) (8)
log M? - EWOH (ULIRGs Only) 21 ?0.49 2.5e?02 ?0.34 3.3e?02 ?0.56 8.4e?03
log M? - EWOH (BAT AGN only) 18 0.04 8.8e?01 0.04 8.2e?01 ?0.00 9.9e?01
log M? - EWOH (Combined Sample) 39 0.13 4.3e?01 0.09 4.0e?01 0.13 4.2e?01
?AGN - EWOH (ULIRGs Only) 37 ?0.42 1.1e?02 ?0.29 1.0e?02 ?0.47 3.7e?03
?AGN - EWOH (BAT AGN only) 38 ?0.11 5.2e?01 ?0.08 5.0e?01 ?0.27 1.0e?01
?AGN - EWOH (Combined Sample) 75 ?0.42 1.8e?04 ?0.29 2.3e?04 ?0.44 7.0e?05
log LAGN - EWOH (ULIRGs Only) 37 ?0.25 1.3e?01 ?0.19 9.7e?02 ?0.31 6.4e?02
log LAGN - EWOH (BAT AGN only) 42 0.03 8.5e?01 0.02 8.6e?01 ?0.03 8.7e?01
log LAGN - EWOH (Combined Sample) 79 0.26 2.0e?02 0.18 2.1e?02 0.25 2.4e?02
Note. ? Column 1: quantities considered for the statistical test. Column 2: number of objects in which OH 119 ?m is
detected. Column 3: Spearman rank order correlation coefficient. Column 4: null probability of the Spearman rank order
correlation coefficient. Column 5: Kendall?s correlation coefficient. Column 6: null probability of Kendall?s correlation.
Column 7: Pearson?s linear correlation coefficient. Column 8: Two-tail area probability of Pearson?s linear correlation. Null
probabilities . 10?3 (shown in bold-faced characters) indicate statistically significant trends.
64
Table 2.7. Results from Statistical Analyses of the Kinematics
Parameter Number of Objects ?s P? ? P? r Pr
(1) (2) (3) (4) (5) (6) (7) (8)
ULIRGs Only
logM? ? v50 18 -0.15 5.6e-01 -0.15 3.8e-01 -0.04 8.7e-01
logM? ? v84 18 -0.10 7.0e-01 -0.08 6.4e-01 0.01 9.6e-01
logSFR? v50 32 0.28 1.2e-01 0.22 8.2e-02 0.06 7.6e-01
logSFR? v84 32 0.19 2.9e-01 0.14 2.4e-01 0.07 7.1e-01
sSFR ? v50 18 0.37 1.3e-01 0.26 1.3e-01 0.54 2.2e-02
sSFR ? v84 18 0.26 3.0e-01 0.21 2.3e-01 0.57 1.4e-02
?AGN ? v50 32 -0.52 2.4e-03 -0.37 3.2e-03 -0.48 5.5e-03
?AGN ? v84 32 -0.46 7.4e-03 -0.31 1.2e-02 -0.46 7.6e-03
logLAGN ? v50 32 -0.50 3.3e-03 -0.37 2.8e-03 -0.56 8.0e-04
logLAGN ? v84 32 -0.44 1.1e-02 -0.32 9.8e-03 -0.54 1.6e-03
BAT AGN Only
logM? ? v50 6 -0.23 6.6e-01 -0.14 7.0e-01 -0.34 5.1e-01
logM? ? v84 6 0.09 8.7e-01 -0.07 8.5e-01 0.27 6.1e-01
logSFR? v50 13 -0.11 7.3e-01 -0.11 6.2e-01 -0.04 8.9e-01
logSFR? v84 13 0.24 4.4e-01 0.17 4.2e-01 -0.02 9.5e-01
sSFR ? v50 6 0.32 5.4e-01 0.28 4.4e-01 0.17 7.4e-01
sSFR ? v84 6 0.54 2.7e-01 0.33 3.5e-01 0.34 5.1e-01
?AGN ? v50 13 -0.50 8.4e-02 -0.32 1.3e-01 -0.50 8.1e-02
?AGN ? v84 13 -0.69 9.4e-03 -0.50 1.7e-02 -0.44 1.3e-01
logLAGN ? v50 14 -0.27 3.6e-01 -0.18 3.7e-01 -0.39 1.7e-01
logLAGN ? v84 14 -0.36 2.1e-01 -0.27 1.9e-01 -0.29 3.2e-01
Combined Sample
logM? ? v50 24 -0.58 2.8e-03 -0.44 2.5e-03 -0.13 5.5e-01
logM? ? v84 24 -0.53 7.9e-03 -0.38 9.9e-03 -0.08 7.0e-01
logSFR? v50 45 -0.34 2.3e-02 -0.21 3.9e-02 -0.44 2.3e-03
logSFR? v84 45 -0.30 4.3e-02 -0.18 7.8e-02 -0.40 6.8e-03
65
Table 2.7 (cont?d)
Parameter Number of Objects ?s P? ? P? r Pr
(1) (2) (3) (4) (5) (6) (7) (8)
sSFR ? v50 24 0.13 5.6e-01 0.10 4.8e-01 0.34 1.1e-01
sSFR ? v84 24 0.10 6.3e-01 0.09 5.2e-01 0.38 6.7e-02
?AGN ? v50 45 -0.04 8.1e-01 -0.03 7.8e-01 -0.08 6.1e-01
?AGN ? v84 45 -0.09 5.6e-01 -0.06 5.7e-01 -0.12 4.5e-01
logLAGN ? v50 46 -0.70 6.8e-08 -0.51 5.7e-07 -0.67 3.3e-07
logLAGN ? v84 46 -0.61 8.2e-06 -0.45 8.9e-06 -0.63 2.9e-06
Note. ? Column 1: quantities considered for the statistical test. Column 2: number of
objects in which OH 119 ?m is detected in either redshifted or blueshifted absorption. Column
3: Spearman rank order correlation coefficient. Column 4: null probability of the Spearman
rank order correlation coefficient. Column 5: Kendall?s correlation coefficient. Column 6: null
probability of Kendall?s correlation. Column 7: Pearson?s linear correlation coefficient. Column
8: Two-tail area probability of Pearson?s linear correlation. Null probabilities . 10?3 (shown
in bold-faced characters) indicate statistically significant trends.
Figure 2.7, Figure 2.8, and Figure 2.9 plot the OH velocities (v50 and v84) as
a function of the stellar masses, star formation rates, and specific star formation
rates. If we consider our two samples individually, a visual inspection of these plots
does not show a correlation between these properties. However, once the samples
are combined we see a negative correlation between the observed OH velocities
and the stellar mass or star formation rate of the galaxy (i.e. galaxies with larger
stellar masses or star formation rates exhibit more negative v50 and v84 values). The
strengths of these correlations are quantified in Table 2.7. No obvious trend is seen
with sSFR for the combined sample. These results remain quantitatively the same
if instead of using the global star formation rates we use the star formation rates
from the central spaxel (scaled by the ratio of the continuum flux from the central
66
400 (a)
0
?400
Abs (BAT AGN)
?800 Abs (ULIRG)
Abs/Emi (ULIRG)
400 (b)
0
?400 NGC 5506
?800
9 10 11 12 13
log(M?/M)
Figure 2.7 v50 and v84 as a function of the stellar masses. The meanings of the
symbols are the same as those in Section 2.5.1.1. The data points joined by a
segment correspond to F14394+5332 W and E. The smaller symbols have larger
uncertainties (values followed by double colons in Table 3.3)
.
spaxal to that of all 25 spaxels, as listed in column (8) of Table 2.2).
Figure 2.10 plots the OH velocities (v50 and v84) as a function of the AGN
fractions ?AGN, where ?AGN is derived from the f30/f15 continuum flux density ratios.
As described in V13, ULIRGs/PG QSOs with dominant AGN (?AGN ? 50%) appear
to have larger negative velocities than ULIRGs/PG QSOs with dominant starbursts
(?AGN ? 50%), but a K-S test between velocity distributions of dominant AGN and
dominant starburst systems indicated no statistically significant difference. A K-S
67
v84 [km s
?1] v ?150 [km s ]
400 (a)
0
?400
Abs (BAT AGN)
Abs/Emi (BAT AGN)
Abs (ULIRG)
?800 Abs/Emi (ULIRG)
400 (b)
0
NGC 7172
?400 IC 5063NGC 5506
NGC 7479
?800
?1 0 1 2 3
log(SFR/[M yr?1])
Figure 2.8 v50 and v84 as a function of the star formation rates. The meanings of the
symbols are the same as those in Section 2.5.1.1. The smaller symbols have larger
uncertainties (values followed by double colons in Table 3.3)
test on the combined BAT AGN and ULIRG + QSO sample also does not show a
significant trend between the measured OH velocity and AGN fraction.
Figure 2.11 shows the OH velocities (v50 and v84) versus the AGN luminosi-
ties, LAGN = 10.5 ? L14?195keV for our BAT AGN sample and LAGN = ?AGNLBOL
for the ULIRG/QSO sample; see Section 2.4. A correlation between these quan-
tities is not observed in the BAT AGN sample, but clear trends are seen in the
ULIRG/QSO sample and in the combined sample (see Table 2.7). Objects with
log (LAGN/L) .11.5 show no evidence for fast outflows. This suggests that AGN
68
v84 [km s
?1] v ?150 [km s ]
400 (a)
NGC 3079 F22491-1808
0
NGC 4102
?400
Abs (BAT AGN)
Abs (ULIRG)
?800 Abs/Emi (ULIRG)
400 (b)
F22491-1808
0 NGC 3079
NGC 4102
?400 NGC 5506
?800
?1200
0 2 4 6
sSFR [Gyr?1]
Figure 2.9 v50 and v84 as a function of the specific star formation rates. The meanings
of the symbols are the same as those in Section 2.5.1.1. The data points joined by
a segment correspond to F14394+5332 E and W. The smaller symbols have larger
uncertainties (values followed by double colons in Table 3.3) The black vertical line
indicates the approximate location of the Main Sequence of star-forming galaxies
(Shimizu et al., 2015).
of lower luminosities are not able to drive significant molecular winds.
69
v84 [km s
?1] v50 [km s?1]
400 (a)
0
?400
?800
400 (b)
0
?400
?800
0 20 40 60 80 100
?AGN[%]
Figure 2.10 v50 and v84 as a function of the AGN fractions. The meanings of the
symbols are the same as those in Section 2.5.1.1. The smaller symbols have larger
uncertainties (values followed by double colons in Table 3.3)
2.7 Discussion
2.7.1 Outflows
2.7.1.1 BAT AGN with Molecular Outflows
As mentioned above, only (one) four objects in our BAT AGN sample shows
(unambiguous) evidence of a molecular outflow (NGC 7479 is the unambiguous case,
while NGC 5506, NGC 7172, and IC 5063 are more uncertain; see Figure 2.8 and
70
v84 [km s
?1] v50 [km s?1]
400 (a)
0
?400
Abs (BAT AGN)
Abs/Emi (BAT AGN)
Abs (ULIRG)
?800 Abs/Emi (ULIRG)
400 (b)
0
NGC 7172
IC 5063
?400 NGC 5506
NGC 7479
?800
9 10 11 12 13
log(LAGN/L)
Figure 2.11 v50 and v84 as a function of the AGN luminosities. The meanings of the
symbols are the same as those in Section 2.5.1.1. The smaller symbols have larger
uncertainties (values followed by double colons in Table 3.3)
Figure 2.11). This corresponds to an outflow detection rate of (6%) 24%, if we take
into account that this search for outflows was possible only in the 17 sources with
OH 119 ?m in absorption. This outflow detection rate is significantly smaller than
that found in ULIRGs (?70%). We note, however, that outflow detection may be
easier in the ULIRGs due to their higher gas fractions.
NGC 7479 is the best case for an outflow in our sample of BAT AGN. NGC
7479 is a barred galaxy with the faint broad line emission at H? but not at H?
of a Seyfert 1.9 galaxy (Ve?ron-Cetty & Ve?ron, 2006). To quantify the uncertainty
71
v84 [km s
?1] v ?150 [km s ]
of the OH blue-wing detection in NGC 7479, we have refit the continuum for this
object with a third order polynomial and then refit the OH doublet via the method
outlined in Section 2.5.1.3. Although we find a shift in the velocities measured with
this different continuum (v50(abs) = ?45 km s?1 and v84(abs) = ?513 km s?1),
this shift (a measure of the error in our velocity estimates) is not large enough to
account for the location of NGC 7479 in Figure 2.11b. The origin for the high outflow
velocity in NGC 7479 may lie in its unusually high far-infrared surface density (Lutz
et al., 2015). The high wind velocity observed in NGC 7479 may thus be due to its
unusually high SFR surface density (e.g. Diamond-Stanic et al., 2012). However, we
cannot exclude the possibility that the OH outflow in NGC 7479 is driven by the
radio jet detected by Laine & Beck (2008).
NGC 5506 is an edge-on disk galaxy optically classified as a Seyfert 1.9 galaxy,
just like NGC 7479. In the very hard X-ray regime, NGC 5506 is one of the most
luminous and brightest Seyferts in the local universe (L14? 43195 keV ? 10 erg s?1;
Baumgartner et al., 2011), and its obscuring column (N = 3.4?1022H cm?2; Bassani
et al., 1999) is intermediate between typical values for Seyfert 1s and 2s. Mid-
infrared 8 ? 13?m observations of NGC 5506 (Roche et al., 1984) suggest that the
nuclear region of NGC 5506 within 5 arcsec is powered by highly obscured AGN
activity with little starburst activity. The OH 119 ?m line profile of NGC 5506
shows a very broad, secondary gaussian component, suggestive of an outflow (see
Figure 2.1), but the velocities of this component are uncertain (see Table 3.3).
Interestingly, the optical forbidden emission lines of NGC 5506 also exhibit distinct
blue wings extending to ?1000 km s?1 (Veilleux, 1991a,b,c), giving credence to the
72
idea that a fast outflow is indeed present in this object.
IC 5063 is a massive (M? ? 1011 M) early-type galaxy which hosts a pow-
erful double-lobed radio source (P 23 ?11.4GHz = 3 ? 10 W Hz ). The presence of this
strong radio source is contrary to the weakly collimated jets found in the BAT AGN
of Middelberg et al. (2004). Observations of high velocity (? 600 km s?1) warm
molecular hydrogen gas in the western lobe of IC 5063 suggests that molecules may
be shock-accelerated by the expanding radio jets (Tadhunter et al., 2014). Thus,
the presence of this powerful jet may explain the additional boost observed in the
(uncertain) outflow velocity v84.
NGC 7172 is an obscured (N 23 ?2H ? 10 cm ; Turner & Pounds, 1989), almost
edge-on Type 2 Seyfert galaxy which belongs to the Hickson compact group HCG
90. Smajic? et al. (2012) derive the two-dimensional velocity field and velocity dis-
tribution of the central 4??? 4?? region of NGC 7172 using several emission lines (i.e.
Pa?, H2(1-0)S(1), and [Si VI]) and two CO stellar absorption features. All mea-
surements indicate disk rotation from east to west with amplitude of at least ?100
km s?1. Comparison of HST F606W imaging (Malkan et al., 1998) with 2MASS
images in J?, H?, and K?bands (Jarrett et al., 2003) shows a shift of the nucleus
by more than 1?? going from the visible to the J?, H?, and K?band. The central
spaxel aperture (9.4??? 9.4??) of Herschel does not have the resolution to detect such
a shift, but if the central spaxel is not centered on the galaxy nucleus, and if the
velocity structure for OH 119 ?m is comparable to that seen in the near-infrared,
then it is possible that the blueshifted absorption we detect is due to incomplete
sampling of the rotation curve rather than a molecular outflow.
73
This last object emphasizes the fact that the poor spatial resolution of Herschel
makes the interpretation of the OH spectra challenging. We have tried to err on
the side of being cautious by utilizing a centroid velocity threshold of ?50 km s?1
and/or the detection of blue wings with v ?184 < ?300 km s . The Herschel data are
not always sensitive to small-scale winds. Indeed, small-scale molecular winds are
known to exist in some of our sources. For instance, Garc??a-Burillo et al. (2014)
detect AGN-driven CO(3-2), CO(6-5), HCN(4-3), HCO+(4-3), and CS(7-6) outflows
in NGC 1068 on spatial scales of ? 20? 35 pc (? 0.3??? 0.5??), while OH is observed
purely in emission in the Herschel data and is therefore ambiguous regarding the
existence of a large-scale outflow until a full analysis of the velocity field is carried
out. Additionally, molecular outflows (or inflows) may be present, but not centered
on the central spaxel. For example, Herschel observations of OH 79, 84, and 65 ?m,
as well as HCN suggest the existence of a spatially resolved outflow outside of the
central spaxel in NGC 3079 even though the central spaxel apparently displays an
OH inflow. A detailed discussion of these objects is outside the scope of this work,
but these issues will be addressed in a future paper.
2.7.1.2 Driving Mechanisms of Molecular Outflows
Correlations between the observed OH outflow velocities and the stellar masses,
star formation rates, and AGN luminosities of the host galaxies can shed light upon
the physical mechanisms responsible for driving molecular winds. There are good
reasons to believe that energy and momentum injection from star formation ac-
74
tivity plays a role in driving massive, galactic-scale outflows. For example, the
observed spatial coincidence between H-? filaments and the X-ray emission from
galactic winds suggests a close physical connection between the origins of these fea-
tures (Cooper et al., 2008). Additionally, previous studies (e.g. Cicone et al., 2014;
Martin, 2005; Rupke et al., 2005a; Schwartz & Martin, 2004; Weiner et al., 2009)
have shown the existence of a positive trend between SFRs and outflow velocities.
Curiously, V13 found no such trend in their ULIRGs + QSO sample. It appears,
as V13 surmised, that the lack of a trend between these quantities was due to the
limited range in SFR of the V13 sample, which spans only ? 2 dex. As seen in
Figure 2.8, when the BAT AGN are combined with the ULIRGs of V13, the sample
spans a range ? 3 dex in SFR, and a positive correlation between outflow velocities
and SFRs becomes apparent.
Positive correlations between stellar masses and outflow velocities in the neu-
tral and ionized phases of the ISM have also been reported in the literature (e.g
Martin, 2005; Rupke et al., 2005a; Weiner et al., 2009). V13 did not see this trend
in the molecular data of their ULIRG + QSO sample, but again attributed this
negative result to the small range in stellar masses of their sample (? 1 dex). Our
BAT AGN + ULIRG + QSO sample now extends over a stellar mass range of ? 2
dex and reveals a significant trend (see Figure 2.7). However, note that this trend
may be of secondary origin since a positive correlation is well known to exist be-
tween stellar mass and SFR in galaxies, both locally (e.g. 0.015 ? z ? 0.1; Elbaz
et al., 2007; Renzini & Peng, 2015; Shimizu et al., 2015) and at high redshift (e.g.
0.7 < z < 1.5; Rubin et al., 2010; Whitaker et al., 2014) ? the so-called Main
75
Sequence of star-forming galaxies.
V13 reported a weak trend between wind velocities (v50 and v84) and AGN
fractions in their ULIRG sample. They cautioned that the correlation could be
merely an obscuration effect where both the AGN and central high-velocity out-
flowing material are more easily detectable when the dusty material has been swept
away or is seen more nearly face-on. Indeed, by adding our BAT AGN sample
to the ULIRG sample, we find that this weak correlation disappears. AGN frac-
tions are thus not a good predictor of molecular outflows. On the other hand,
a convincing causal connection was presented in V13 between AGN luminosities
and molecular outflow velocities with a possible steepening of the relation above
log(LbreakAGN /L) = 11.8 ? 0.3. Our results do indeed strongly suggest that at higher
AGN luminosities (log(LAGN/L) & 11.5) the AGN dominates over star formation
in driving the outflow. At lower luminosities (log(LAGN/L) . 11.5) the AGN may
not have the energetics required to drive fast molecular winds.
Statistically, the correlation between wind velocity and AGN luminosity is
stronger than the correlation between wind velocity and SFR (see Table 2.7). How-
ever, while the AGN may be the dominant source for driving the winds observed
in our sample, it is clear that SFR also plays a role in driving these winds. The
presence of an AGN seems to boost the observed velocity over that which would be
observed in purely star forming systems (Cicone et al., 2014), although large scatter
is observed (see Figure 2.11). This scatter may have multiple origins: (1) If the
outflow is not spherically symmetric, projection effects will produce scatter in the
observed velocities. (2) The efficiency to entrain material in the outflow depends on
76
several complex factors associated with the acceleration mechanisms and the multi-
phase nature of these processes. For instance, if radiation pressure is the dominant
mechanism driving the outflow, radiative transfer effects are probably important so
that not all of the OH-detected gas is ?seeing? the same radiation field and there-
fore not experiencing the same radiation force. Similarly, if the dominant driving
mechanism is ram pressure of a fast diffuse medium on dense cloudlets, one might
expect to observe a range of velocities depending on the characteristics (e.g., sizes
and densities) of the cloudlets entrained in the flow and their survival timescales
(e.g. Cooper et al., 2008, 2009). (3) The AGN luminosity is measured from the 14
? 195 keV flux and therefore represent the current value of the AGN luminosity. In
contrast, the observed outflow was likely produced ? 1 to several ?106 years ago
based on the measured OH velocities and inferred outflow sizes (? 1 kpc; Gonza?lez-
Alfonso et al. (2014b, 2015); Sturm et al. (2011)). AGN variability may therefore
cause considerable scatter in Figure 2.11.
2.7.2 Inflows
Seven objects (Centaurus A, Circinus, NGC 1125, NGC 3079, NGC 3281,
NGC 4945, and NGC 7582) have OH absorption features with median velocities
larger than 50 km s?1, corresponding to an inflow detection rate of ?40%. By
far the best case for inflow in our sample is Circinus where OH 119 ?m shows an
inverted P-Cygni profile. Inverted P-Cygni profiles are also tentatively detected in
Centaurus A and NGC 3281.
77
Interestingly, previous searches for neutral-gas (Na I D) outflows/inflows in
IR-faint Seyfert galaxies have also shown distinctly higher detection rates of inflows
than outflows. Specifically, in an analysis of 35 IR-faint Seyferts (109.9 < LIR/L <
1011.2), Krug et al. (2010) reported an inflow detection rate of ? 37% and an outflow
detection rate of only ? 11%. These numbers are similar to those derived here, and
considerably different from those measured in (U)LIRGs using OH (? 10%; V13)
and Na I D (e.g. ? 15% inflow detection rate for 78 starbursting galaxies with
log(LIR/L) = 10.2 ? 12.0; Rupke et al., 2005a,b). The origin of this difference is
unknown. The fast winds in (U)LIRGs may disturb the neutral-molecular gas and
prevent it from infalling to the center. On average, IR-bright sources are also richer
in gas and dust than IR-faint galaxies. This material may be masking the central
regions where inflow is taking place.
2.7.3 The 9.7 ?m Silicate Feature
The analysis of the strength of the S9.7?m feature can provide insight into the
mechanism responsible for the excitation of OH 119 ?m observed in our sample. As
seen in Figure 2.5, the comparison of the OH 119 ?m equivalent width, EWOH, and
the strength of the silicate feature, S9.7?m , implies a rather tight connection between
OH gas and mid-infrared obscuration (a correlation is also found between OH 65
?m and S9.7?m ; see also Gonza?lez-Alfonso et al. (2015)). The clear trend of deeper
OH 119 ?m absorption and fainter OH 119 ?m emission with increasing silicate
obscuration (more positive values of S9.7?m ) suggests that OH 119 ?m emission is
78
strongly affected by the obscuring geometry. Our results expand on those of V13
and S13 who found a similar trend with OH equivalent width and the strength of
the 9.7 ?m feature among ULIRGs.
S13 argue that the OH 119 ?m emission region often lies within the buried
nucleus and that radiative excitation is the dominant source of OH 119 ?m emission.
In reality, the geometry of the silicate obscuration and the source of the OH 119
?m emission may be more complex. Figure 2.12 plots the total equivalent width of
OH 119 ?m as a function of S9.7?m and distinguishes between AGN spectral type for
each object. Note that objects classified as LINERs have been excluded from this
plot due to the ambiguous energy source in these objects (Sturm et al., 2006). None
of the Type 1 galaxies (BAT AGN or ULIRGs) show a strong silicate absorption
(S9.7?m & 1.5). Interestingly, we see that OH 119 ?m for some Type 2 BAT AGN
and ULIRGs is observed in emission (EWOH < 0 km s
?1) while the silicate feature,
S9.7?m , is seen in absorption. It is possible that the OH 119 ?m emission region is
not nuclear, but is instead distributed throughout a circumnuclear starburst where
the number density (n(H2) ? a few times 105 cm?3), temperature (? 100 K), and
OH abundance (X(OH) ? 2?10?6) are sufficient for collisional rather than radiative
excitation of the upper level of OH 119 ?m (e.g. NGC 1068; Spinoglio et al., 2005).
Until we examine in detail all 5 ? 5 spaxels of the PACS data, the location of the
OH emission in our objects will remain unclear. This type of analysis can be done
in just a few select objects; this will be discussed in a future paper.
Determining the origin of the silicate feature is also a challenge. The dust
responsible for this feature may reside in the AGN torus, or on larger scale in the
79
300
200
100
0
BAT AGN Type 1
? BAT AGN Type 2100 ULIRG Type 1
ULIRG Type 2
?1 0 1 2 3 4 5
S9.7?m
Figure 2.12 Total (absorption + emission) equivalent widths of OH 119 ?m as a
function of the apparent strength of the 9.7 ?m silicate feature relative to the local
mid-infrared continuum. Note that S9.7?m is a logarithmic quantity and can be
interpreted as the apparent silicate optical depth. The strength of this absorption
feature increases to the right. Also note that objects classified as LINERs have
been excluded from this plot. Filled markers refer to Type 1 and open markers
refer to Type 2. Blue squares and red circles represent BAT AGN and ULIRGs/PG
QSOs, respectively. Vertical lines represent objects in which OH was not detected.
Horizontal lines represent objects with a null S9.7?m detection. Dotted lines and
dash-dotted lines refer to Type 1 and Type 2, respectively. Blue and red lines
indicate BAT AGN and ULIRGs/PG QSOs, respectively.
disk of the host galaxy, or some combination of these two.
Figure 2.13 plots the ratio of the semi-minor and semi-major axes (a proxy for
the inclination of the host galaxy disk) as a function of S9.7?m for BAT AGN. We
have excluded the ULIRG/PG QSO sample from this particular analysis because
these objects are undergoing or have undergone a major merger, and therefore, do
not have a well-defined galactic plane or inclination. Visual inspection of Figure 2.13
suggests a weak trend between the inclination of the BAT AGN host galaxy and
the depth of the 9.7 ?m silicate feature. First, we see that OH for nearly face-on
80
EW ?1OH [km s ]
1.0 Abs (BAT AGN) Type 1
Abs (BAT AGN) Type 2
Emi (BAT AGN) Type 1
Emi (BAT AGN) Type 2
Abs/Emi (BAT AGN) Type 1
Abs/Emi (BAT AGN) Type 2
Null (BAT AGN) Type 1
Null (BAT AGN) Type 2
0.5
NGC 4945
0.?0 1 0 1 2 3 4 5 6
S9.7?m
Figure 2.13 Ratio of the semi-minor axis to the semi-major axis (a proxy for the
inclination of the host galaxy disk) as a function of S9.7?m for the BAT AGN sam-
ple. Squares, triangles, and circles represent BAT AGN in which OH is observed
purely in absorption, purely in emission, composite absorption/emission, respec-
tively. Diamonds represent objects in which OH was undetected. Filled points and
dash-dotted lines indicate Type 1 while open points and dotted lines indicate Type 2
AGN. Horizontal lines represent objects with a null 9.7 ?m silicate feature detection.
hosts (b/a & 0.8) is seen only in pure emission and the strength of S9.7?m is weak.
Second, we find that nearly edge-on galaxies show the broadest range of silicate
absorption strengths, including the most extreme case of NGC 4945. However, the
lack of a strong trend between inclination and S9.7?m strength suggests that the dust
responsible for the 9.7 ?m silicate feature is located not only in the plane of the host
galaxy, but also in the nuclear torus. Our results are qualitatively consistent with
Goulding et al. (2012), who invoke a clumpy torus paradigm of many individual
optically thick clouds (Nenkova et al., 2002, 2008) and suggest that deeper silicate
features (S9.7?m & 0.5) are due to dust distributed at radii much larger ( pc) than
that predicted for a torus.
81
b/a
2.8 Conclusions
We present the results of our analysis of Herschel/PACS spectroscopic obser-
vations of 52 nearby (d < 50 Mpc) BAT AGN selected from the very hard X-ray
(14-195 keV) Swift-BAT Survey of local AGN. We also include in our analysis the
Herschel/PACS data on 38 ULIRGs and 5 PG QSOs from V13. The depth of
the silicate feature at 9.7 ?m feature in all these objects is measured from archival
Spitzer/IRS data. Our combined BAT AGN + ULIRG + QSO sample covers a
range of AGN luminosity and SFR of & 3 dex and ? 3 dex, respectively. The main
results from our analysis are:
1. The OH 119 ?m feature is detected in 42 of the 52 objects in our BAT AGN
sample. Of these detections, OH 119 ?m is observed purely in emission for
25 targets, purely in absorption for 12 targets, and absorption+emission com-
posite for 5 targets.
2. Evidence for molecular outflows (absorption profiles with median velocities
more blueshifted than ?50 km s?1 and/or blueshifted wings with 84-percentile
velocities less than ?300 km s?1) is seen in only four objects (NGC 7479, NGC
5506, NGC 7172, and IC 5063), corresponding to 24% of all targets where this
search for outflows was possible. This outflow detection rate is significantly
smaller than that found in ULIRGs by V13 (?70%). The best case for outflow
is NGC 7479, an object with one of the highest infrared surface densities among
our BAT AGN.
82
3. We find evidence for molecular inflows (absorption profiles with median veloc-
ities more redshifted than 50 km s?1) in seven objects (Circinus, NGC 1125,
NGC 3079, NGC 3281, and NGC 4945), corresponding to an inflow detection
rate of ?40%, considerably higher than the rate measured among ULIRGs
(?10%) but similar to the detection rate of neutral-gas (Na I D) inflows among
IR-faint Seyfert galaxies. By far the best case for OH inflow among our BAT
AGN is Circinus, where OH 119 ?m shows a distinct inverted P-Cygni profile.
4. The positive correlation between OH velocities and AGN luminosities reported
in V13 is strengthened in the combined sample, but it is not seen in the BAT
AGN sample. This suggests that luminous AGN play a dominant role in
driving the fastest winds, but stellar processes will dominate when the AGN
is weak or absent.
5. We confirm that the absorption strength of OH 119 ?m is a good proxy for
dust optical depth in these systems. Our findings are consistent with most,
but not all, of the OH 119 ?m emission originating near the AGN. Spatially
resolved OH emission is seen in a few objects in our sample (e.g., NGC 1068)
and may originate from a circumnuclear starburst where the number density
and temperature are sufficiently high to collisionally excite the upper level of
OH 119 ?m.
6. A comparison of the strength of the 9.7 ?m silicate feature with the inclination
of the host galaxy disk and spectral type of the AGN confirm earlier results
that the dust responsible for this feature is located not only in the nuclear
83
component, but also in the disk of the host galaxy.
84
Chapter 3: Constraints on the OH-to-H Abundance Ratio in Infrared-
Bright Galaxies Derived from the Strength of the OH 35 ?m
Absorption Feature
3.1 Introduction
Galactic-scale outflows are driven by the collective effects of supernovae, winds
from massive stars, and winds or jets from active galactic nuclei (AGN; e.g. Veilleux
et al., 2005, and references therein). These outflows can disperse heavy elements,
interstellar dust, and energy throughout a galaxy and into the galaxy halo; they may
therefore play a significant (if not dominant) role in galaxy evolution (e.g. Benson
et al., 2003; Bower et al., 2006; Croton et al., 2006; Fabian, 2012; Hopkins et al.,
2008, 2006; King, 2010; Springel, 2005). They have been detected ubiquitously in
both low- and high-redshift systems (e.g. Alexander et al., 2010; Banerji et al., 2011;
Brusa et al., 2016; Cano-D??az et al., 2012; Carniani et al., 2015, 2016; Cicone et al.,
2015; Cresci et al., 2015; Fo?rster Schreiber et al., 2014; Genzel et al., 2014; Greene
et al., 2012; Harrison et al., 2014, 2012, 2016; Kornei et al., 2012; Maiolino et al.,
2012; Martin et al., 2012; McElroy et al., 2015; Nesvadba et al., 2011; Tremonti et al.,
2007; Veilleux et al., 2005; Weiner et al., 2009; Zakamska et al., 2016). They are
85
particularly significant in infrared luminous galaxy mergers where both starbursts
and luminous AGN often co-exist together (e.g. Cicone et al., 2014; Gonza?lez-Alfonso
et al., 2017; Janssen et al., 2016; Martin, 2005, 2006; Rupke et al., 2017, 2002,
2005a,b,c; Rupke & Veilleux, 2011; Spoon et al., 2013; Sturm et al., 2011; Veilleux
et al., 2017, 2013). However, the central regions of these mergers are generally
enshrouded in dust so infrared tracers are needed to peer through the dusty veil.
Numerous studies have demonstrated that the infrared transitions of the hy-
droxyl molecule, OH, are a powerful diagnostic tool with which to investigate the
nature of cool molecular outflows. Fischer et al. (2010) first reported the remarkable
detection of OH 79 and 119 ?m (hereafter, the OH ??doublet transitions will be
referred for short only by their rounded wavelengths: OH 79 and OH 119) with P-
Cygni profiles (incontrovertible evidence of the presence of an outflow) in the nearby
QSO, Mrk 231. Results from Sturm et al. (2011) revealed the detection of massive
molecular outflows traced by OH 79 for six galaxies and a tentative correlation be-
tween OH outflow velocities and AGN luminosities. This correlation was confirmed
by Veilleux et al. (2013) (and Spoon et al., 2013) who reported the detection of OH
outflows in ?70% of 43 (24) ULIRGs + QSOs, and further strengthened when Stone
et al. (2016) combined the 43 ULIRGs + QSOs of Veilleux et al. (2013) with OH
119 observations of 52 Swift/Burst Alert Telescope (BAT)-detected AGN.
In order to evaluate the impact of these outflows on the evolution of their
host galaxies, it is necessary to measure the sizes and energetics (e.g., mass-outflow
rate, momentum flux, and mechanical power) of these outflows. Multi-transitional
OH analyses have had considerable success constraining these quantities and the
86
dominant mechanisms driving these winds (Gonza?lez-Alfonso et al., 2014b, 2017).
All of the dynamical quantities depend on the hydrogen column densities inferred
from the OH features. These in turn depend on the OH?to?H abundance ratio,
XOH, needed to infer the H column densities from the measured OH column den-
sities. This abundance ratio is generally assumed to be ?2.5 ? 10?6, based on the
value inferred from multitransition observations of OH in the Galactic Sgr B2, the
Orion KL outflow, and in buried nuclei, as well as predictions of chemical models
of photodissociation regions and of cosmic-ray- and X-ray-dominated regions (see
Gonza?lez-Alfonso et al., 2017, and references therein).
In this paper, we provide an independent constraint on XOH. We focus our at-
tention on the OH 35 ?m (hereafter, OH 35) absorption feature covered by Spitzer-
IRS. OH 35 is optically thinner than the other OH doublets in the far-infrared, and
although it is more sensitive to extinction effects, OH 35 remains a reliable indicator
of OH column densities. OH is assumed to be mixed with dust and given the high
continuum optical depths (even in the far-IR; Gonza?lez-Alfonso et al., 2014b), the
absorption is restricted to a ?skin? of matter. Our OH column densities therefore
also provide a constraint on the abundance of OH relative to the dust, from which
XOH (relative to H nuclei) is derived, assuming a gas?to?dust ratio. We are able
to carry out this analysis on 15 local (z ? 0.06) (ultra-)luminous infrared galaxies
(U/LIRGs), which have both Herschel-PACS observations of the OH 119 and/or 79
?m transitions as well as Spitzer-IRS observations of the OH 35?m transition.
Background information on the structure of the OH molecule is presented in
Section 3.2. The galaxy sample, data reduction, and spectral analysis are described
87
Table 3.1. Spitzer -IRS Spectra: Observations
Name PI AORKey(s)
(1) (2) (3)
Arp 220 Houck, James 16910080, 21247232
IRAS F04454-4838 Armus, Lee 20334080
IRAS F05189-2524 Calibration, IRS 04969216
IRAS F08572+3915 Calibration, IRS 20925696, 21457920,
24568064, 27381760,
28704256
IRAS F12224-0624 Bradford, Charles 11272960
IRAS 15250+3609 Houck, James 04983040
IRAS 17208-0014 Houck, James 04986624
IRAS F20551-4250 Houck, James 04990208
IRAS F23128-5919 Houck, James 04991744
Mrk 231 Calibration, IRS 16694016, 17102592,
17202688, 19318016,
21453568, 22157312,
24915456, 34294016
Mrk 273 Houck, James 04980224
NGC 2623 Houck, James 09072896
NGC 4418 Houck, James 04935168
UGC 5101 Houck, James 04973056
Zw453.062 Sturm, Eckhard 10512128
Note. ? Column 1: Galaxy name. Column 2: Principle Investi-
gator. Column 3: AORKey.
in Section 3.3. Section 3.4 presents the results of the OH 35 line profile fitting.
Section 3.5 discusses the implications of our results. A summary of the results and
conclusions is given in Section 4.10.
88
Table 3.2. Galaxy Properties
Name Other Name za f15/f30 ?AGN logLbol logLSB logLAGN ?? MH Type
[%] [L] [L ] [L ] [km s
?1
  ] mag
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
IRAS F04454-4838 0.0529 0.09b 42.0 11.84c 11.61 11.47 ? ? ? ? ? ? ? ? ?
IRAS F05189-2524 0.0426 0.198 71.7 12.22 11.67 12.07 137 ?23.96 AGN 2
IRAS F08354+2555 NGC 2623 0.01846 0.07b 29.6 11.58c 11.05 11.43 ? ? ? ? ? ? L
IRAS F08572+3915 0.0584 0.191 70.4 12.2 11.67 12.05 ? ? ? ?23.58 L
IRAS F09320+6134 UGC 5101 0.0394 0.13 56.4 12.05 11.69 11.8 ? ? ? ? ? ? L
IRAS F12224-0624 0.02636 0.04b 0 11.34c ? ? ? 11.34 ? ? ? ? ? ? L
IRAS F12243-0036 NGC 4418 0.00727 0.128 55.7 11.06 10.71 10.81 ? ? ? ? ? ? AGN 2
IRAS F12540+5708 Mrk 231 0.0422 0.272 80.5 12.6 11.89 12.51 120 ?24.52 AGN 1
IRAS F13428+5608 Mrk 273 0.0378 0.081 34.2 12.21 12.03 11.74 285 ?24.32 AGN 2
IRAS 15250+3609 0.0554 0.095 42.3 12.1 11.86 11.73 150 ? ? ? L
IRAS F15327+2340 Arp 220 0.0181 0.049 5.8 12.22 12.19 10.98 164 ? ? ? L
IRAS 17208-0014 0.0428 0.038 < 5.0 12.45 12.43 < 11.15 229 ? ? ? L
IRAS F20551-4250 0.043 0.132 56.9 12.11 11.74 11.87 140 ? ? ? L
IRAS F23024+1916 Zw453.062 0.02510 0.07b 30.0 11.36c 10.84 11.21 ? ? ? ? ? ? L
IRAS F23128-5919 ESO 148-IG 002 0.0446 0.154 63 12.09 11.66 11.89 ? ? ? ? ? ? H
Note. ? Column 1: Galaxy name. Column 2: Another name. Column 3: Redshift. Column 4: f15/f30 values from Veilleux et al.
(2009). Column 5: ?AGN, fractional contribution of the AGN to the bolometric luminosity based on the f15/f30 ratio. Column 6: Bolometric
luminosity. Column 7: Starburst bolometric luminosity. Column 8: AGN luminosity. Column 9: ??, stellar velocity dispersion from Dasyra
et al. (2006a,b, 2007). Column 10: MH , absolute H?band magnitudes from Veilleux (2006); Veilleux et al. (2009). Column 11: Optical
spectral types, H for H II galaxies, L for LINER?like, AGN 2 and AGN 1 for type 2 AGN and type 1 AGN, respectively. For the spectral
type, we adopted the classification from Veilleux et al. (1999, 2009). When not available, we used the values from NED/SIMBAD.
aNASA/IPAC Extragalactic Database (NED); https://ned.ipac.caltech.edu/
bf15/f30 ratio is measured from low-resolution Spitzer -IRS spectra extracted from CASSIS.
cValues for L(40 ? 500?m) and DL (the luminosity distance) used to calculate L(8 ? 1000?m) and Lbol (Sanders & Mirabel, 1996) are
adopted from Armus et al. (2009)
89
3.2 Background: The OH Molecule
The OH electronic ground state 2? is split due to spin-orbit coupling where
spin angular momentum can be oriented either parallel or anti?parallel to the orbital
angular momentum. Thus, the rotational states are split into two ladders: 2?3/2
and 2?1/2 (Storey et al., 1981), where the total angular momentum is given by
J = N ? 1/2 (N is the rotational quantum number). Additionally, the orbital
angular momentum can have two orientations with respect to the molecular axis,
each having a slightly different energy. This ?-doubling further splits each J level
into two levels of opposite parity.
The cross-ladder 2? ? 23/2 ?1/2 J = 3/2? 5/2 ?-doublet transition at 34.60
and 34.63 ?m (see Fig.1 of Gonza?lez-Alfonso et al., 2014b) is an efficacious column
density estimator. Within an optical depth region of ?35?m ? 0.5, we expect OH
35 to remain optically thin, enabling us to calculate the OH abundance for a given
gas-to-dust ratio. Indeed, the spontaneous transition probabilities (described by the
Einstein?A coefficient) of the intra-ladder transitions are much larger (? 0.1 s?1 for
J = 3/2 to ? 10 s ?1 for J = 13/2) than the cross-ladder transitions (? 0.05? 10?4
s?1) from the same levels (Offer & van Dishoeck, 1992). Excitation of the cross-
ladder transitions at 35 ?m decay through multiple branches (i.e. 35 ?m; 15-84-119
?m; 48-119 ?m; 99-53 ?m; 99-163-79 ?m). Only 4% of all 35 ?m absorption events
result in re-emission at 35 ?m (Bradford et al., 1999). Thus, OH 35 can provide a
meaningful constraint on the true column density.
Note that observations of OH 35 have also proved useful in studies of OH
90
megamasers (Darling, 2007; Darling & Giovanelli, 2006). Radiative pumping from
the absorption of 35 ?m photons can invert the hyperfine population of the OH
ground state. A 18 cm continuum source can then stimulate maser emission. Evi-
dence in support of this radiative pump model was reported in Skinner et al. (1997)
who observed OH 35 in Arp 220. They determined that the photon absorption rate
in the 35 ?m transition could be efficient enough to power observed OH masers.
3.3 Sample Selection, Data Reduction, and Spectral Analysis
3.3.1 The Sample
We have searched the Herschel Science Archive1 (HSA) and the Spitzer Her-
itage Archive2 (SHA) for objects with (1) z . 0.06 (the OH 35 feature of more
distant objects will be redshifted out of the Spitzer -IRS wavelength range), (2)
PACS spectra covering the redshifted OH 79 and/or 119 ?m features, and (3) OH
35 ?m detection in the IRS long high (LH) spectra. There were 242 objects which
met the first two criteria and only 15 of those objects satisfied all three criteria.
Details of the Spitzer observations for our sample are listed in Table 3.1 and galaxy
properties may be found in Table 3.2.
All sources are morphologically classified as mergers or interacting systems
(e.g. Veilleux et al., 2002). Our sample exhibits a range of interaction stages, from
double nuclei systems with projected separation < 10 kpc (e.g. Mrk 273, IRAS
F14348?1447, IRAS F20100?4156, IRAS F08572+3915), to mergers showing a sin-
1http://www.cosmos.esa.int/web/herschel/science-archive
2http://sha.ipac.caltech.edu/applications/Spitzer/SHA/
91
gle nucleus with tidal tails (e.g. IRAS F05189?2524), to warm, quasar-dominated
late stage mergers (e.g. Mrk 231).
The bolometric luminosities of the galaxies are similar (log(Lbol/L) ? 12.2),
but the AGN contribution to the bolometric luminosity (i.e. the AGN fraction)
essentially runs the full gamut from 0 to 1, resulting in AGN luminosities in the
range of log(LAGN/L) = 10.81? 12.51. The AGN fractions here are adopted from
Veilleux et al. (2009), which are based on the 30-to-15 ?m continuum flux ratio,
f30/f15.
3.3.2 Data Reduction
All of the mid-infrared (5 ? 37 ?m), ?high resolution? (R ? 600) spectra used
for this study were obtained with the Long High (LH) module of the Infrared Spec-
trograph (IRS; Houck et al., 2004) on board the Spitzer Space Telescope (Werner
et al., 2004). These spectra were acquired using a high-accuracy blue peak-up and
observed in staring mode. The spectra were extracted using either the S17.2 or
the S18.7 pipeline of the Spitzer Science Center (SSC). Data reduction follows the
procedure outlined in Spoon et al. (2009) and in Spoon & Holt (2009), and starts at
the DROOP level products created by the SSC pipeline. These products have been
corrected for stray-light contributions, non-linear responsivity in the pixels, and
drooping, but have not been flat fielded. For some observations, sky background
images were then subtracted, and interpolation over bad pixels was performed us-
92
ing the IDL package IRSCLEAN1. The data analysis package SMART (Higdon et al.,
2004) was then used to extract the one-dimensional target spectrum in ?full slit?
(11.1?? ? 22.3??) mode.
The same procedure was followed for a large number of observations of the
calibration star ? Dra. The relative spectral response function (RSRF) was then
created from the ratio of the observed spectrum of ? Dra to the stellar reference
spectrum (Cohen et al., 2003, Sloan et al. 2005, private communication). The
final spectrum was obtained by multiplying the object spectrum by the RSRF. For
objects with more than one AOR Key, the final spectrum is the result of co-adding
(no special weighting applied) the high-resolution spectra from each AOR key.
For most of our objects, there is an artifact at? 34.8?35.0?m in the observer?s
frame in the NOD 1 spectrum, but the artifact is not present in the NOD 2 spectrum.
Inspection of the RSRF for NOD 1 shows that there is a 20% dip in the response
at these wavelengths. It is possible that the dip in the NOD 1 spectrum therefore
originates in the RSRF. The vertical light-green band in Figure 3.1 shows the region
in velocity space where the dip occurs. Fortunately, its location does not affect the
OH 35 and [Si II] spectral line structures. The wavelength calibration of our spectra
is accurate to about 100 km s?1 (e.g. about 1/5 of an IRS resolution element of
500 km s?1).
1IRSCLEAN is available at http://irsa.ipac.caltech.edu/data/SPITZER/
docs/dataanalysistools/tools/irsclean/
93
18 [S III]
IRAS 17208-0014
16
k= 3
14
k= 2
12
k= 1
10
OH 35 [Si II]
?12000 ?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000
Velocity [km s?1]
Figure 3.1 Example of twelve different continuum fits to the mid-infrared spectrum
of IRAS F17208-0014. The black solid histograms are the data separated from each
other by an arbitrary offset. Spline pivot points are anchored at wavelengths of 33.0
?m, 33.8 ?m, 34.5 ?m, and 35.0 ?m. The blue, green, and red dots indicate the
set of pivot points which are respectively the lower limits (v1), upper limits (v2),
and best-fit-by-eye values (v3) used to fit a continuum spline. The top most set of
fits is the result of using a third-order (k = 3) spline fit to each pivot point set.
Below are the results for second-order (k = 2) and first-order (k = 1) spline fits. In
each case, the blue, green, and red dotted lines show the results of the spline fits
to the lower-limit, upper-limit, and ?best-fit-by-eye? continuum flux pivot points.
The last spectrum at the bottom is fit with a polynomial. The magenta regions in
this spectrum indicate the regions used to fit the first (solid blue line), second (solid
green line), and third (solid red line) order polynomials to the continuum. The
two dotted vertical lines in grey mark the rest wavelengths of the OH 35 doublet at
systemic velocity. The dotted vertical lines in red mark the locations of the emission
features [S III] 33.48 ?m and [Si II] 34.82 ?m. The vertical light-green band shows
the region where a 20% dip in the detection response in the NOD 1 position occurs.
94
Flux Density [Jy]
[S III] 3.4 [S III] 9.5 [S III]
46
Arp 220 IRAS F04454-4838 IRAS F05189-2524
3.2
44 9.0
3.0
42
8.5
40 2.8
8.0
38 2.6
36 2.4 7.5
OH 35 [Si II] OH 35 [Si II] OH 35 [Si II]
34 2.2
7.0
?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
2.3
6.50 [S III] [S III] 5.50 [S III]
IRAS F08572+3915 2.2 IRAS F12224-0624 IRAS 15250+3609
6.25 5.25
2.1
6.00
5.00
2.0
5.75
1.9 4.75
5.50
1.8 4.50
5.25
1.7 4.25
5.00
4.75 OH 35 [Si II]
1.6 OH 35 [Si II] 4.00 OH 35 [Si II]
1.5
?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
8.0
[S III] [S III] [S III]
13 IRAS 17208-0014 IRAS F20551-4250 5.5 IRAS F23128-5919
7.5
12 5.0
7.0
11 4.5
6.5
10 4.0
6.0
9 OH 35 [Si II] OH 35 [Si II] 3.5 OH 35 [Si II]
5.5
?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
10.0
12
24 [S III] [S III] [S III]
Mrk 231 Mrk 273 9.5 NGC 2623
23
11 9.0
22
8.5
21
10
20 8.0
19
9 7.5
18
OH 35 [Si II] OH 35 [Si II] 7.0 OH 35 [Si II]
17 8
?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
33 5.0 4.5[S III] [S III] [S III]
NGC 4418 UGC 5101 Zw453.062
32
4.0
4.5
31
3.5
30
4.0
3.0
29
28 3.5 2.5
27
2.0
26 OH 35 [Si II] 3.0 OH 35 [Si II] OH 35 [Si II]
1.5
?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000 ?12000?10000 ?8000 ?6000 ?4000 ?2000 0 2000 4000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
Figure 3.2 Fits to the 33?35?m continua for all 15 objects in the sample. The grey
shaded area shows the full range of the twelve different continuum fits described in
Figure 3.1. Black dots are median values of the six fluxes at each pivot point, and
the dotted line is the resultant third-order spline fit to those pivot points. The grey
dotted vertical lines mark the rest wavelengths of the OH 35 doublet at systemic
velocity. The red dotted vertical lines mark the locations of the emission features [S
III] at 33.48 ?m and [Si II] at 34.815 ?m. The vertical light-green band shows the
region where a 20% dip in the detection response in the NOD 1 position occurs.
95
Flux Density [Jy] Flux Density [Jy] Flux Density [Jy] Flux Density [Jy] Flux Density [Jy]
Flux Density [Jy] Flux Density [Jy] Flux Density [Jy] Flux Density [Jy] Flux Density [Jy]
Flux Density [Jy] Flux Density [Jy] Flux Density [Jy] Flux Density [Jy] Flux Density [Jy]
3.3.3 Spectral Analysis
To determine the extent to which continuum placement affects the measured
properties (e.g. total integrated flux and equivalent width) of the OH 35 feature,
we have fit twelve different baselines to the spectra. First-, second-, and third-order
polynomials were fit via least squares minimization using wavelength regions which
avoided spectral features (e.g. emission or absorption lines) and known artifacts.
Splines were also fit to the continuum via four pivot points anchored at wavelengths
of 33.0 ?m, 33.8 ?m, 34.5 ?m, and 35.0 ?m. Pivot point wavelengths were chosen
to avoid known spectral features within this mid-infrared region.
Three sets of continuum fluxes for the pivot points were chosen by eye. One
set marks an upper limit on the continuum fluxes while another marks a lower limit
on the continuum fluxes. The third set is a ?best fit? by visual inspection. To each
set of pivot points, a first-, second-, and third-order spline was fit. Figure 3.1 shows
an example of the continuum fits to IRAS F17208-0014.
Figure 3.2 shows the range of continuum placements (grey shaded area) for
each object. The black dots are median values of the six fluxes at each pivot point,
and the dotted line is the resultant third-order spline fit to those pivot points. We
choose this ?median? fit as the continuum for our sources and subtract it from each
spectrum. The dotted vertical lines in red mark the locations of emission features
[S III] at 33.48 ?m and [Si II] at 34.815 ?m. For both Figure 3.1 and Figure 3.2,
the two dotted vertical lines in grey mark the rest wavelengths for both components
of the OH 35 doublet at systemic velocity. These are separated by only 0.03 ?m or
96
? 250 km s?1. Given the spectral resolution of the data (?500 km s?1), a single
blended spectral feature is detected in our sources.
Line profile fits of the OH 35 ??doublet were computed by using PySpecKit, a
spectroscopic analysis and reduction toolkit for optical, infrared, and radio spectra
(Ginsburg & Mirocha, 2011). The toolkit uses the Levenberg-Marquardt technique
to solve the least-squares problem in order to find the best fit for the observa-
tions. Profile fitting of the OH doublet followed a similar procedure as that outlined
in Veilleux et al. (2013) and in Stone et al. (2016), in which the line profile was
modeled using two Gaussian components, each characterized by their negative am-
plitude, peak position, and standard deviation (or, equivalently FWHM). Gaussian
parameters of the components were allowed to vary independently of each other.
Note that these fits are only used to quantify the strength (i.e. total integrated flux
and equivalent width) of the OH 35 absorption feature. The Gaussian components
do not have a physical interpretation, nor do they account for each unresolved line
in the doublet.
3.4 Results
Figure 3.3 shows the fits to the continuum-subtracted OH 35 feature of each
galaxy. The dashed blue lines indicate the Gaussian components and the magenta
line is their sum. The total integrated fluxes and equivalent widths of the OH
35 feature measured from these spectral fits and their associated 1-? uncertainties
based on the fits are listed in Table 3.3.
97
0.6
2 0.1Arp 220 IRAS F04454-4838 0.4 IRAS F05189-2524 IRAS F08572+3915
0.4
0.0 0.2
0 0.2
0.0
?0.1
?2 ?0.2 0.0
?0.2
?0.4
? ?0.24 ?0.3
?0.6
?0.4
?6 ?0.4 ?0.8
?0.6
?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
0.6 1.5
0.1 IRAS F12224-0624 0.4 IRAS 15250+3609 IRAS 17208-0014 IRAS F20551-42501.0
0.5
0.0 0.2 0.5
? 0.0
0.0
0.1 0.0
?0.2 ?0.5
?0.2
? ?1.0 ?0.50.4
?0.3
? ?1.50.6
?0.4 ?1.0
?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
1.0
1.5
1.5
1.5 IRAS F23128-5919 Mrk 231 Mrk 273 NGC 2623
0.5
1.0 1.0
1.0
0.0 0.5
0.5
0.5
?0.5 0.0
0.0
0.0 ? ?0.51.0
?0.5
?1.0
?0.5 ?1.5
?1.5 ?1.0
?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
6 1.5 2.5
NGC 4418 UGC 5101 2.0 Zw453.062
4
1.0
1.5
2
1.0
0.5
0 0.5
0.0 0.0?2
?0.5
?4 ?0.5 ?1.0
?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
Figure 3.3 Two-Gaussian fits to the continuum-subtracted OH 35 line profiles of the
15 objects in our sample; see Section 3.3.3. In each figure, the solid black histogram
is the data. Blue dashed lines indicate the two Gaussian components which best fit
the line profile, and the magenta line is the sum of those two components. The grey
dotted vertical lines mark the rest wavelengths of the OH 35 doublet at systemic
velocity. The red dotted vertical line marks the location of the [Si II] emission line
at 34.815 ?m.
98
Flux Density [Jy] Flux Density [Jy] Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy] Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy] Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy] Flux Density [Jy]
8 8 8 8
IRAS 17208-0014
6 6 6 6
p3 v1k3 v2k3 v3k3
4 4 4 4
2 p2 2 v1k2 2 v2k2 2 v3k2
0 0 0 0
p1 v1k1 v2k1 v3k1
?2 ?2 ?2 ?2
?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
Figure 3.4 Examples of OH 35 line profile fits for each of the twelve different con-
tinuum subtracted spectra in IRAS F17208-0014. Line colors and styles are the
same as that for Figure 3.3. ?p1, p2, p3? indicate first-, second-, and third-order
polynomial fits to the continuum, respectively. ?v1? refers to the ?lower-limit? pivot
points in Figure 3.1. ?v2? refers to the ?upper-limit? pivot points, and ?v3? to the
?best-fit-by-eye? pivot points. ?k1, k2, k3? are the orders of the spline fitted to the
continuum. For example, ?v2k3? is the third-order spline fit to the continuum using
the ?upper-limit? pivot points.
Our sources show the OH 35 line profile in absorption only. This is to be
expected since essentially all OH molecules pumped to the upper 2?1/2, J = 5/2
level will relax by emitting a 99 ?m photon along the 2?1/2 ladder instead of emitting
a 35 ?m photon. We note that this 99 ?m emission line is not observed in the PACS
observations of our sample because it is redshifted into the 100 ?m gap of PACS.
Figure 3.4 shows the twelve OH 35 line profile fits to each of the continuum-
subtracted spectra of IRAS F17208-0014. ?p1, p2, p3? indicate first-, second-, and
third-order polynomial fits to the continuum, respectively. ?v1? refers to the lower
limit continuum flux pivot points in Figure 3.1. ?v2? refers to the upper limit
continuum flux pivot points, and ?v3? refers to the ?best fit by eye? pivot points.
?k1, k2, k3? are the orders of the spline fitted to the continuum. For example,
?v2k3? is the third order spline fit to the continuum using the upper limit pivot
points. The resultant fluxes and equivalent widths from all of the continuum fits
99
Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy]
Flux Density [Jy]
Integrated Flux
(a)
8
6
4
2
0
0 500 1000 1500 2000 2500 3000
[Jy km s?1]
Wvel
(b)
6
4
2
0
0 25 50 75 100 125
[km s?1]
Figure 3.5 Distributions of the measured OH 35 (a) integrated fluxes and (b) equiv-
alent widths.
are used to estimate the measurement errors. Section 3.4 shows the distributions of
the measured OH 35 fluxes and equivalent widths in our sample.
On average, we find that the uncertainties on the placement of the continuum
translate into uncertainties of about ?15% on the final values of the equivalent width
measurements listed in Table 3.3. Since the OH column density scales linearly with
the OH 35 equivalent width, the typical uncertainties on the measurements of NOH
and XOH are also ?15%. We note that the detection of OH 35 in UGC 05101
is marginal (with an uncertainty as high as ?64%), so it is excluded from these
calculations and from the mean OH column density and mean OH?to?H abundance
ratio in the following sections.
100
Table 3.3. OH 35 ?m Profile Properties
Name Flux Wvel NOH XOH
[Jy km s?1] [km s?1] [1017 cm?2] [10?6]
(1) (2) (3) (4) (5)
Arp 220 2977 ? 362 72 ? 9 1.69 ? 0.21 1.64 ? 0.20
IRAS F04454-4838 190 ? 15 64 ? 6 1.49 ? 0.15 1.44 ? 0.14
IRAS F05189-2524 524 ? 60 61 ? 7 1.42 ? 0.17 1.38 ? 0.16
IRAS F08572+3915 176 ? 22 30 ? 4 0.7 ? 0.09 0.68 ? 0.10
IRAS F12224-0624 109 ? 7 54 ? 4 1.26 ? 0.08 1.22 ? 0.08
IRAS 15250+3609 262 ? 23 52 ? 5 1.21 ? 0.12 1.17 ? 0.11
IRAS 17208-0014 1286 ? 292 108 ? 27 2.51 ? 0.63 2.44 ? 0.61
IRAS F20551-4250 633 ? 132 90 ? 20 2.09 ? 0.46 2.03 ? 0.45
IRAS F23128-5919 249 ? 27 60 ? 7 1.39 ? 0.16 1.35 ? 0.16
Mrk 231 1050 ? 186 51 ? 10 1.19 ? 0.22 1.15 ? 0.22
Mrk 273 531 ? 118 51 ? 12 1.19 ? 0.28 1.15 ? 0.28
NGC 2623 300 ? 9 36 ? 1 0.86 ? 0.02 0.83 ? 0.03
NGC 4418 1581 ? 403 53 ? 14 1.23 ? 0.32 1.20 ? 0.32
UGC 5101 67 ? 38 18 ? 11 0.42 ? 0.26 0.41 ? 0.26
Zw453.062 93 ? 8 42 ? 4 1 ? 0.08 0.95 ? 0.08
Note. ? Column 1: Galaxy name. Column 2: Total integrated flux. Col-
umn 3: Equivalent width. Column 4: OH column density. Column 5: OH
abundance relative to H.
101
log(N ?2OH/[cm ])
(a)
6
4
2
0
16.50 16.75 17.00 17.25 17.50
log(XOH)
(b)
6
4
2
0
?6.50 ?6.25 ?6.00 ?5.75 ?5.50
Figure 3.6 Distributions of the (a) OH column densities and (b) OH-to-H abundance
ratios, XOH, inferred from the fits of the OH 35 feature.
3.4.1 OH Column Density
The equivalent width of an absorption line is defined as
?
Iv(c)? Iv
Wv = dv, (3.1)
Iv(c)
where Iv(c) is the intensity of the continuum and Iv is the line intensity at frequency
v. The optical depth is related to the line intensity via
Iv = Iv(c) e
??v , (3.2)
102
?
where ? = n? ds (n is the volume density of absorbers and ? is the cross section
for absorption). Therefore, we can relate the optical depth and the equivalent width,
?
Wv = (1? e??v) dv. (3.3)
For a foreground screen of homogeneous, isothermal gas illuminated by a back-
ground source, in the optically thin limit (?  1), Equation 3.3 reduces to
?
W? = NOH ?0 ?? d?, (3.4)
where NOH is the ?ground state OH column density and ?? is the line profile function?
defined such that ?? d? = 1 and ?0 holds the parameters of the transition. If the0
Einstein coefficients Au` for spontaneous emission and Bu` for stimulated emission
are in units of [s?1] and [cm3 (J s2)?1], respectively (where u refers to the upper
energy level and ` the lower energy level), we have the relations
8? h
Au` = B3 u`, (3.5)?
Bu` gu = B`u g`, (3.6)
where g is the degeneracy of the level and h is the Planck constant. Thus,
h A ?2u` gu
?0 = B`u = . (3.7)
? 8? g`
Substituting ?0 in Equation 3.4 and noting that Wvel [cm s
?1] = W [s?1? ] ?? [cm],
103
we see that the number of molecules per unit area along the line of sight is
Wvel8? g`
NOH = . (3.8)
A ?3u` gu
Adopting Au` = 0.0174 s
?1 from Destombes et al. (1977) and Bradford et al. (1999),
we calculate OH column densities using the empirically measured equivalent widths
for our objects. Those column densities are shown in Figure 3.6 and listed in Ta-
ble 3.3 with the associated 1-? standard deviation of the distribution. We find a
mean of NOH = 1.31? 0.22? 1017 cm?2 for the objects in our sample.
3.4.2 XOH
We can now calculate XOH, the OH abundance relative to H
NOH
XOH = . (3.9)
NH
First, we estimate the column density of H nuclei for a given ?35?m (the optical depth
at 35 ?m due to dust) via
(Mgas/Mdust) ?35?m
NH = , (3.10)
?35?mmH
where (Mgas/Mdust) is the gas-to-dust ratio by mass, ?35?m is the mass absorption
coefficient of dust at 35 ?m, and mH is the H mass. From Figure 2 of Gonza?lez-
Alfonso et al. (2014a) we adopt ? 2 ?135?m = 290 cm g , and we adopt a gas-to-dust
104
ratio of 100, appropriate for the nuclear regions of luminous and ultraluminous
infrared galaxies (e.g. Wilson et al., 2008). We adopt an optical depth ?35?m =
0.5 because, in the inner regions, the OH excitation temperature will be in near
equilibrium with the dust temperature. Moreover, for ?35?m > 0.5, OH absorption
becomes harder to detect due to extinction at 35?m. Thus, we find NH ? 1? 1023
cm?2.
Results for the OH column density from Equation 3.8 combined with the H
column density from Equation 3.10 provide XOH. We calculate this abundance ratio
for each of our sources (see Table 3.3) with the associated 1-? standard deviation
of the distribution. We find a mean of XOH = 1.27 ? 0.21 ? 10?6 for our sample.
Figure 3.6 shows the OH-to-H abundance ratios resulting from these calculations.
We remark that our estimates for XOH are formally only lower limits because
(1) some fraction of the OH molecules are in excited levels, so NOH is a lower limit
to the column density of all OH states, (2) the covering fraction of the 35 ?m
continuum source by this obscuring material may not be 100%, and (3) some of the
spectra are not background-subtracted which reduces the equivalent width of the
OH 35 feature1
1When comparing the optimally extracted spectra available from CASSIS with the spectra
presented here, we find that the background subtraction does not have a noticeable effect on the
equivalent widths.
105
3.5 Discussion
3.5.1 XOH: Comparison with the Literature
Here we compare our empirically derived values of XOH with those found in the
literature. Estimates of XOH are typically derived by fitting radiative transfer mod-
els to multitransition observations of OH. For example, Goicoechea & Cernicharo
(2002) modeled the first 20 rotational levels in OH and constrained the model with
observations of Sgr B2. They report a range of X ?6OH = 2? 5? 10 (OH relative to
H2). A similar study by Goicoechea et al. (2006) of the Orion KL outflow derived a
lower value of XOH = 0.5 ? 1 ? 10?6. Following these two papers, the OH outflow
studies of Sturm et al. (2011) and Gonza?lez-Alfonso et al. (2014b, 2017) adopted a
value of XOH = 2.5 ? 10?6. Recent estimates of XOH in obscured nuclei (Falstad
et al., 2015; Fischer et al., 2014; Gonza?lez-Alfonso et al., 2012) seem to favor the
higher OH abundances. Our empirically derived value of XOH = 1.3 ? 0.4 ? 10?6
falls in the lower portion of this range of values. However, recall that our estimate is
a lower limit, so it is formally consistent with all previously assumed and computed
values of XOH in the literature.
Our estimate is also consistent with Richings & Faucher-Giguere (2017)?s broad
range of values for XOH (3.3?10?6 ? 2.2 ? 10?5), which is deduced from hydro-
chemical simulations of AGN-driven winds. It should be noted, however, that the
XOH values here and in Gonza?lez-Alfonso et al. (2014b, 2017) are relative to hydro-
gen nuclei in both atomic and molecular forms, while Richings & Faucher-Giguere
106
(2017)?s abundance ratios are relative to molecular hydrogen, hence the larger val-
ues. Finally, we note that our empirically determined values of XOH are consistent
with the values of chemical models of photodissociation regions (the peak value
in Sternberg & Dalgarno, 1995) and of cosmic-ray- and X-ray-dominated regions
(Meijerink et al., 2011).
3.5.2 XOH: A Check on Radiative Transfer Models
In recent years, radiative transfer models have been used to infer the main
physical parameters of molecular outflows in several infrared-bright galaxies (e.g.
Gonza?lez-Alfonso & Cernicharo, 1999; Gonza?lez-Alfonso et al., 2014b, 2017), and
many of these properties depend upon the inferred H column density (e.g. outflowing
mass, mass outflow rate, momentum flux, and energy flux). It is therefore important
to verify that these models also predict realistic OH 35.
Figure 3.7 shows the predicted OH 35 feature of the five ULIRGs (IRAS
F05189-2524, IRAS F08572+3915, IRAS F20551-4250, Mrk 231, and Mrk 273) in
common with the sample of Gonza?lez-Alfonso et al. (2017). In this paper, two
ground-state lines (OH 79 and 119) and two excited lines (OH 65 and 84) were
used to constrain the model. The green line is the OH 35 profile predicted by these
models while the magenta line is the result from our fits to the observed feature (see
Section 3.3.3). The agreement between the two is generally very good in terms of
overall strength (equivalent width) of the feature. The predicted profiles are also
largely consistent with the observed profiles, after taking into account the rather
107
1.15 1.15 1.15
IRAS F05189-2524 IRAS F08572+3915 IRAS F20551-4250
1.10 1.10 1.10
1.05 1.05 1.05
1.00 1.00 1.00
0.95 0.95 0.95
0.90 0.90 0.90
?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1] Velocity [km s?1]
1.15 1.15
Mrk 231 Mrk 273
1.10 1.10
1.05 1.05
1.00 1.00
0.95 0.95
0.90 0.90
?3000 ?2000 ?1000 0 1000 2000 3000 ?3000 ?2000 ?1000 0 1000 2000 3000
Velocity [km s?1] Velocity [km s?1]
Figure 3.7 OH 35 spectra normalized to the continuum. Red and blue lines are the
two Gaussian components of the fitted line and the magenta line is the sum of the
two components. The green line is the profile predicted by the radiative transfer
model described in Gonza?lez-Alfonso & Cernicharo (1999); Gonza?lez-Alfonso et al.
(2017). The modeled line profiles are constrained by observations of four OH lines
at 65, 79, 84, and 119 ?m.
large uncertainties on the observed profiles associated with the continuum place-
ment and the modest spectral resolution of the IRS data. The consistency between
the model predictions and the observations adds some confidence in the energetics
derived from these models.
108
3.6 Summary
We have analyzed mid-infrared (5 ? 37 ?m) Spitzer?IRS spectra of 15 nearby
(z . 0.06) ULIRGs focusing on the OH 35 ?m absorption feature. Several dif-
ferent methods were used to fit these spectra and estimate the uncertainties on
the measured line parameters. The measured equivalent widths of OH 35 im-
plies an average OH column density and 1-? standard deviation to the mean of
NOH = 1.31 ? 0.22 ? 1017 cm?2. We infer an OH-to-H abundance ratio of X =
1.27 ? 0.21 ? 10?6 when we assume a gas-to-dust ratio of 100, appropriate for the
nuclear regions of luminous and ultraluminous infrared galaxies. This abundance
ratio is formally a lower limit. It is consistent to within a factor of two with the
value assumed in the radiative transfer models of OH outflows in these systems
(XOH = 2.5 ? 10?6). We also find that the depths and profiles of OH 35 predicted
by these models are largely consistent with the observed OH 35 spectra, providing
support for the predicted energetics from these models.
109
Chapter 4: Far-Infrared Integral-Field Spectroscopy of Nearby
Galactic Winds with Herschel-PACS
4.1 Introduction
Large-scale galactic outflows driven by stellar processes and/or active galactic
nuclei (AGNs) play a pivotal role in the life cycle of galaxies by heating the cold inter-
stellar medium (ISM), suppressing or triggering star formation, fueling or quenching
accretion onto the central black hole, and by influencing the chemical enrichment
of galaxies (e.g. Fabian, 2012; Heckman & Thompson, 2017; Rupke, 2018; Veilleux
et al., 2005, 2020). Although we now have a better understanding of how winds may
significantly impact the formation and evolution of galaxies (e.g. we see this in semi-
analytic, N-body, and hydrodynamical models of galaxy evolution where feedback is
incorporated, (e.g. Naab & Ostriker, 2017; Somerville & Dave?, 2015; Zhang, 2018),
there still exists a plethora of unanswered questions concerning the detailed physics
and properties of galactic outflows. What is the primary mechanism of energy and
momentum injection that powers an outflow? How can we constrain outflow prop-
erties such as morphologies, lifetimes, spatial extent, radial-dependent velocities,
mass outflow rates, and energetics? What is the efficiency of entraining material
110
from the ISM and ejecting it out of the disk? The answers to these questions can
fundamentally measure the impact of outflows on galaxy evolution. However, it is
challenging to constrain these properties because outflows involve multiple phases
of the ISM (cold molecular, cool neutral atomic, warm and hot ionized; Heckman
& Thompson, 2017) and hence, span a large range of temperatures (T ? 10 ? 107
K; Zhang, 2018), physical scales (a few parsecs to several kiloparsecs), and densities
(n ? 10?2 ? 106 cm?3; Croxall et al., 2012; Draine, 2011).
If we wish to gain a complete understanding of how winds contribute to the
evolution of galaxies, it is crucial that we conduct detailed multi-phase ISM investi-
gations of outflows and we must look to our local neighborhood where nearby galax-
ies provide us the opportunity to study spatially resolved spectroscopic observations
of outflows. Integral Field Spectroscopy (IFS) of the ionized (e.g. Harrison et al.,
2014; Martin & Soto, 2016; Venturi et al., 2017), neutral (e.g. Rupke & Veilleux,
2011), and molecular (e.g. Pereira-Santaella et al., 2016; Rupke & Veilleux, 2013)
ISM phases of outflows has made it possible to decompose the complex kinematics
within galaxies. Not only can we utilize IFS to delineate disk rotation and outflow
in a galaxy, but we can also gain insight into the spatial extent or morphological
structure of the outflow.
However, star formation regions and areas near the AGN in a galaxy are heav-
ily enshrouded in dust which severely hampers UV and optical wavelength obser-
vations. Fortunately, the far-infrared (FIR) provides us the opportunity to observe
in these dusty regions without large extinction effects. We take advantage of the
unprecedented combination of angular resolution, sensitivity, and spatial coverage of
111
the FIR spectroscopic integral field unit (IFU) of the Photodetector Array Camera
and Spectrometer (Poglitsch et al., 2010, PACS;) on board the Herschel Space Ob-
servatory (Pilbratt et al., 2010) to peer through the dusty veil obscuring the regions
where stars form and black holes grow.
In this paper, we present the main results of our analysis of the cool neu-
tral atomic and the warm ionized ISM phases, as traced by FIR fine-structure line
transitions in seven nearby (d < 16 Mpc) galaxies that are well known to harbor
winds: Centaurus A (hereafter abbreviated as Cen A), Circinus, M 82, NGC 253,
NGC 1068, NGC 3079, and NGC 4945. The proximity of these galaxies allows us
to spatially resolve the morphologies and kinematics of the outflows (at the dis-
tances for our sample, the physical sizes covered by a PACS 9.7??? 9.7?? spaxel are
? 0.2 ? 1 kpc). The OH 119 ?m doublet (hereafter, OH 119) has also been ob-
served by PACS in this galaxy sample, which we present and compare with the
observations of the ionized and neutral gas.
The PACS data of the atomic gas for M 82 have been examined in detail in
Contursi et al. (2013, hereafter C13). However, their observations cover a larger field
of view (FOV), 2.5??2.5?, compared to our ?47??? 47?? FOV. Therefore, they were
able to capture the outflow out to 1 kpc above and below the galaxy disk whereas
we only capture the outflow out to a few hundred pc. Moreover, the majority of
our FOV falls inside the starburst region of M 82. The observations from C13 were
obtained in mapping mode instead of the pointed mode spectra presented here.
Due to the spatial steps of the rasters used in their observations, the sky is Nyquist
sampled and a spatial resolution of 6?? (? 130?270 pc, depending upon wavelength)
112
is achieved. In pointed mode, the beam is undersampled and has a native resolution
of ? 10??. This is important to note because the sky will be undersampled which
results in an underestimated source flux and the true morphology of an object will be
degraded and difficult to reconstruct from a single pointed observation. Fortunately,
we can increase confidence in the approximation in which the source morphology
is reconstructed by comparing the interpolated observation with higher resolution
data. Therefore, we re-examine PACS data of M 82 here for completeness and to
validate our analysis on the other objects.
This paper is organized as follows: The FIR atomic fine-structure lines are
discussed in Section 4.2. The galaxies in our sample are described in Section 4.3.
Descriptions of the observations, archival data, and data reductions are found in
Section 4.4. We present the results of our analysis in Section 4.5 and Section 4.6.
Section 4.7 and Section 4.8 outline the methods used to derive the gas outflow
properties. The interpretation of the results and their implications are discussed in
Section 4.9. Section 4.10 outlines our main conclusions. Appendix B contains the
analysis of the molecular gas traced by OH 119. In this paper we adopt the standard
convention that negative velocities with respect to systemic indicate approaching
material.
4.2 Atomic Fine-Structure Emission Lines
The gas phases in the PACS FIR range (55?210 ?m) cool through strong emis-
sion lines arising from atomic fine-structure transitions and molecular transitions,
113
with the strongest atomic emission originating from carbon, nitrogen, and oxygen
and their various ions. These atoms are collisionally excited and then de-excite
through forbidden transitions, emitting photons and thus removing thermal energy
from the gas. Indeed, the cooling radiation from these emission lines are so powerful
that they provide almost all of the gas cooling in the cold neutral and warm ionized
phases of the ISM, and hence the emission is easily detectable over large distances
(for example, [O III] 88 has been observed out to z = 9.11; Hashimoto et al., 2018).
The importance of these fine-structure transitions cannot be understated since they
are the dominant mechanism for gas cooling and can therefore directly impact the
star formation in a galaxy. Here we observe five atomic fine-structure line transi-
tions within the PACS FIR waveband: [O I] 63 ?m, [O III] 88 ?m, [N II] 122 ?m,
[O I] 145 ?m, and [C II] 158 ?m (hereafter, we drop the wavelength unit, ?m, for
brevity). Because this suite of FIR lines span a wide range of ionization potentials
(IP = 11.26 ? 35.12 eV), critical densities (ncrit ? 40 ? 105 cm?3), and excitation
temperatures (Tex = 91? 326 K), the combination of these lines allows us to inves-
tigate the physical conditions of the multiple phases of an outflow and probe the
different regions of a galaxy from which these emission lines originate (see Table 4.1
for more details of the fine-structure lines).
The low-excitation atomic lines ([O I] 63 , [O I] 145 , [C II] 158 ) sample pho-
todissociation regions (PDRs) at the interfaces between molecular, atomic, and
ionized gas phases in star-forming regions where intense far-ultra-violet (FUV)
radiation photodissociates CO, resulting in bright emission of [O I] and [C II]
114
Table 4.1. Fine-structure Lines
Line Transition ? ? IP Tex ncrit,H ncrit,e
[?m] [GHz] [eV] [K] [cm?3] [cm?3]
(1) (2) (3) (4) (5) (6) (7) (8)
[O I] 63 3P ?31 P2 63.18 4744.77 0.0? 13.62 227 2? 105 ? ? ?
[O III] 88 3P1?3P0 88.36 3393.01 35.12? 54.94 163 ? ? ? 510
[N II] 122 3P2?3P1 121.89 2459.38 14.53? 29.60 188 ? ? ? 375
[O I] 145 3P 30? P1 145.52 2060.07 0.0? 13.62 326 5? 104 ? ? ?
[C II] 158 2P3/2?2P1/2 157.74 1900.54 11.26? 24.38 91 2? 103 47
Note. ? The information in Table 4.1 is based on Herrera-Camus et al. (2018a), Israel
et al. (2017), Farrah et al. (2013) and the Leiden Atomic and Molecular Database (Scho?ier
et al., 2005). Column 1: Atomic fine-structure line, Column 2: Electronic energy level
transition, Column 3: Rest wavelength, Column 4: Rest frequency, Column 5: Ionization
potential to create the species and the ionization potential to ionize the species, Column
6: Excitation temperature, Column 7: Critical density for collisions with H I and H2,
Column 8: Critical density for collisions with electrons.
(Hollenbach & Tielens, 1997; Sternberg & Dalgarno, 1995; Tielens & Hollenbach,
1985). [C II] 158 arises from both ionized and neutral gas due to its ionization
potential of 11.2 eV. [C II] 158 is the dominant coolant in regions with densities
nH? 10 ? 105 cm?3 and temperatures T ? 100 ? 300 K, and is the strongest
emission line from cooler gas (T < 104 K) in galaxies (Carilli & Walter, 2013).
In PDRs, the [O I] 63 and [O I] 145 line emission originates solely in the neutral
regime. [O I] 145 is the faintest of the fine-structure lines, but its advantage is that
it does not suffer from self-absorption like [O I] 63, which can be affected by self-
absorption by relatively small amounts of foreground gas (N ? 2 ? 1020 cm?2H ;
Liseau et al., 2006). [O I] 63 line emission can also originate in X-ray dominated
regions (XDRs) near AGN (Dale et al., 2004). At higher densities (ne ? 105 cm?3)
and higher temperatures (T > 200 K), [O I] 63 becomes the dominant cooling line
115
for neutral atomic gas instead of [C II] 158 (Meijerink et al., 2007). [O I] 63 line
emission may also arise from shocks (Lutz et al., 2003).
The high-excitation ionic emission line [O III] 88 traces moderate density (?
100 cm?3) clouds, H II regions of photoionization from young, massive stars (Spinoglio
& Malkan, 1992; Voit, 1992), and the narrow line region (NLR) excited by AGN
(D??az-Santos et al., 2017). The high abundance of oxygen makes [O III] 88 a pri-
mary coolant of the warm ionized phase of the ISM, and its high ionization potential
(35.12 eV) predominantly traces very energetic conditions near AGN (Ferkinhoff
et al., 2010), hot OB stars, or shocks (Stasin?ska et al., 2015). [N II] 122 , a low-
excitation line, also originates in H II regions and is solely excited in ionized gas
due to its ionization potential of 14.53 eV, a value just above the ionization energy
of hydrogen, 13.6 eV. Moreover, its low critical density (? 350 cm?3) means that
it can easily be excited in the diffuse ionized phase of the ISM (Goldsmith et al.,
2015; Herrera-Camus et al., 2016).
4.3 The Sample
The seven galaxies in our sample (see Table 4.2 for more details about the
galaxy sample) are well known to harbor winds and have been studied intensively
across the entire accessible electromagnetic spectrum. The spatial scale of these
studies span a range of a few pc to several hundred kpc. Two of our sample are
classified as pure starbursts (?H II galaxies? in the table; M 82 and NGC 253) while
the other five (Cen A, Circinus, NGC 1068, NGC 3079, and NGC 4945) harbor both
116
a narrow-line AGN and a starburst.
M 82 is considered to be the archetypal starburst galaxy, hosting a substantial
population of supernova remnants (Pedlar et al., 1999) and several star clusters (de
Grijs, 2001). It is the energy from these populations? stellar processes that drives
the well-known kpc-scale bipolar conically-shaped outflow along the galaxy?s minor
axis. The inclination of M82 provides a direct sight line to the S part of the disk
which makes the approaching S outflow cone clearer in visibility. The N outflow cone
is receding from us and lies behind the galaxy disk where emission in the base of the
outflow may be extinguished by the disk itself. Hard X-ray (E ? 6.7 keV) emission
inside the central |r| < 200 pc, |z| < 100 pc of the galaxy arises from the wind
fluid that drives the larger scale outflow (Griffiths et al., 2000). UV (e.g. Hoopes
et al., 2005), mid-infrared (Engelbracht et al., 2006), and far-infrared/sub-mm (e.g.
Alton et al., 1999; Roussel et al., 2010) images provide evidence of dust entrained in
the hotter outflowing gas. Confined to the cone walls, warm ionized gas is seen as
optical filamentary emission in H?, [N II], and [S II] (Shopbell & Bland-Hawthorn,
1998). Warm molecular hydrogen in the near-infrared roughly traces the H? and
warm dust emission (Veilleux et al., 2009). Observations of CO show molecular
line splitting along the minor axis providing further evidence of a conical outflow
morphology (Leroy et al., 2015).
117
Table 4.2. Galaxy Properties
Galaxy Name RA Dec Distance SFR log LFIR i PA Type Morph.
[h m s] [? ? ??] [Mpc] [M ?1 ? ? yr ] [L] [ ] [ ]
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
M 82 09 55 52.2 +69 40 46.60 3.6 10 10.67 81 65 SB I0
Cen A 13 25 27.62 ?43 01 08.81 3.8 2 9.91 75 278 AGN & SB S0 pec
Circinus 14 13 09.95 ?65 20 11.87 4.2 4.7 9.92 65 216 AGN & SB SAb
NGC 253 00 47 33.12 ?25 17 17.59 3.9 3 10.35 76 230 SB SABc
NGC 1068 02 42 40.71 ?00 00 47.81 14.4 18 11.13 40 278 AGN & SB S0 pec
NGC 3079 10 01 57.80 +55 40 47.24 16 10.64 84 166 AGN & SB SBc
NGC 4945 13 05 27.48 ?49 28 05.57 3.8 0.4 10.28 75 45 AGN & SB SBcd
Note. ? Column 1: Galaxy name, Column 2: Adopted RA of the galaxy center, Column 3: Adopted Dec of
the galaxy center, Column 4: Luminosity distance, Column 5: Star formation rate, Column 6: FIR luminosity,
Column 7: Galaxy inclination, Column 8: Position angle of the galaxy major axis, Column 9: Type, Column
10: Morphology classification.
118
Cen A is the prototypical Fanaroff-Riley I-type low-luminostiy galaxy, and at
a distance of 3.8 Mpc (Harris et al., 2010), it is the nearest active radio galaxy. In
the radio, Cen A hosts a collimated subparsec-scale jet and counterjet (Burns et al.,
1983; Tingay et al., 1998), parsec-scale inner lobes (Clarke et al., 1992; Schreier et al.,
1981), a kpc-scale middle lobe (Morganti et al., 1999), and giant outer radio lobes
on scale of hundreds of kpc (McKinley et al., 2013). Optical images show evidence
of a bipolar outflow on scales of ? 10 kpc (McKinley et al., 2018). H? (Blanco et al.,
1975; McKinley et al., 2018) and [O III] ?5007 filaments are also seen on scales of
the middle lobe.
Circinus is the closest Seyfert 2 galaxy at 4.2 Mpc (Tully et al., 2009) and
the second brightest AGN in the mid-infrared (after NGC 1068). The prominent
ionization cone located NW of the galaxy nucleus contains extended optical filaments
with outflowing velocities ? 100 ? 200 km s?1 out to ? 1 kpc (e.g. Marconi et al.,
1994; Sharp & Bland-Hawthorn, 2010; Veilleux & Bland-Hawthorn, 1997; Veilleux
et al., 2003). The ionization cone and outflow are traced on parsec scale in the NIR
(Maiolino et al., 2000) and on kiloparsec scale in the X-Ray (Mingo et al., 2012).
CO observations detect a ?NW cloud? spatially coincident with the ionized outflow
cone with a blue-shifted velocity of ? 150 km s?1 and a ?far W cloud? moving along
an H? filament (Zschaechner et al., 2016). Edge-brightened bipolar lobes are also
seen in the radio which extend up to more than 2 kpc on either side of the galaxy
plane (Elmouttie et al., 1998a).
NGC 253 is a starburst-dominated galaxy at a distance of 3.9 Mpc (West-
moquette et al., 2011), making it one of the two closest nuclear starburst galaxies
119
(M 82 being the other). Chandra observations (Strickland et al., 2002; Strickland
& Stevens, 2000) show a limb-brightened conical outflow to the SE of the galaxy
nucleus which is spatially aligned with the ?100?300 km s?1 outflow observed in H?
(Westmoquette et al., 2011), and a northern X-ray outflow lobe that lies behind the
disk of the galaxy. Outflows have been observed in Na I D (Heckman et al., 2000)
and in OH (Sturm et al., 2011; Turner, 1985). A prominent molecular wind has
been traced in CO with a mass outflow rate between ? 3 ? 9 Myr?1 (assuming
optically thin CO emission; Bolatto et al., 2013) and ? 25? 50 M yr?1 (assuming
optically thick CO emission; Krieger et al., 2019; Zschaechner et al., 2018). Even at
the lower limit, these outflow rates (when compared to NGC 253?s star formation
rate) are large enough to substantially affect the galaxy?s star formation activity.
NGC 1068 is the archetypal Type 2 Seyfert galaxy and contains an AGN
and a circumnuclear starburst. NGC 1068 also harbors a kiloparsec-scale radio jet
(e.g. Wilson & Ulvestad, 1987). The NLR gas has been observed as an outflow with
a biconical morphology (Cecil et al., 1990; Das et al., 2006). The northeast side of
the cone is in front of the galactic disk (oriented towards us) while the Southwest
side is behind (oriented away from us) and obscured by the galactic disk (Barbosa
et al., 2014). Sub-mm interferometry of molecular lines in the circumnuclear disk
strongly suggests the existence of a giant, AGN-driven outflow extending to ? 100
pc with a velocity of ? 100 ? 200 km s?1 (Garc??a-Burillo et al., 2014; Krips et al.,
2011).
120
Table 4.3. PACS Fine-Structure Line Observations
Object Name ObsId Line Spec. Res.a Ang. Res.b,c t dexp AOT Chop Throw Program ID
(1) (2) (3) (4) (5) (6) (7) (8) (9)
M 82 1342186799 [O I] 63 90 9.5 582 line large SDP esturm 3
M 82 1342186798 [O III] 88 125 9.5 986 line large SDP esturm 3
M 82 1342186798 [N II] 122 290 10.0 986 line large SDP esturm 3
M 82 1342186798 [O I] 145 250 11.0 986 line large SDP esturm 3
M 82 1342186798 [C II] 158 240 11.5 986 line large SDP esturm 3
Cen A 1342202590 [O I] 63 90 9.5 873 line large KPGT esturm 1
Cen A 1342202588 [O III] 88 125 9.5 11291 range med KPGT rguesten 1
Cen A 1342203444 [N II] 122 290 10.0 5663 range large KPGT rguesten 1
Cen A 1342202589 [O I] 145 250 11.0 2341 line large KPGT esturm 1
Cen A 1342202589 [C II] 158 240 11.5 2341 line large KPGT esturm 1
Circinus 1342191298 [O I] 63 90 9.5 1281 line large KPGT esturm 1
Circinus 1342191297 [O III] 88 125 9.5 1948 line large KPGT esturm 1
Circinus 1342191297 [N II] 122 290 10.0 1948 line large KPGT esturm 1
Circinus 1342191297 [O I] 145 250 11.0 1948 line large KPGT esturm 1
Circinus 1342191297 [C II] 158 240 11.5 1948 line large KPGT esturm 1
NGC 1068 1342191153 [O I] 63 90 9.5 3671 range large KPGT esturm 1
NGC 1068 1342203124 [O III] 88 125 9.5 3386 range large KPGT esturm 1
NGC 1068 1342191154 [N II] 122 290 10.0 3944 range large KPGT esturm 1
NGC 1068 1342191154 [O I] 145 250 11.0 3944 range large KPGT esturm 1
NGC 1068 1342203121 [C II] 158 240 11.5 3456 range large KPGT esturm 1
NGC 253 1342199414 [O I] 63 90 9.5 873 line large KPGT esturm 1
NGC 253 1342199415 [O III] 88 125 9.5 1975 line large KPGT esturm 1
NGC 253 1342199415 [N II] 122 290 10.0 1975 line large KPGT esturm 1
NGC 253 1342199415 [O I] 145 250 11.0 1975 line large KPGT esturm 1
NGC 253 1342199415 [C II] 158 240 11.5 1975 line large KPGT esturm 1
121
Table 4.3 (cont?d)
Object Name ObsId Line Spec. Res.a Ang. Res.b,c t dexp AOT Chop Throw Program ID
(1) (2) (3) (4) (5) (6) (7) (8) (9)
NGC 3079 1342221391 [C II] 158 240 11.5 8045 range med DDT esturm 4
NGC 4945 1342212218 [O I] 63 90 9.5 2939 range large KPGT esturm 1
NGC 4945 1342212220 [O III] 88 125 9.5 1975 line large KPGT esturm 1
NGC 4945 1342212220 [N II] 122 290 10.0 1975 line large KPGT esturm 1
NGC 4945 1342212220 [O I] 145 250 11.0 1975 line large KPGT esturm 1
NGC 4945 1342212220 [C II] 158 240 11.5 1975 line large KPGT esturm 1
aSpectral resolutions are in units of km s?1 .
bAngular resolutions are in units of arcsec.
cEstimated from Fig.8 in the PACS Spectroscopy Performance and Calibration Guide Issue 3.0.
dExposure times are in units of seconds.
Note. ? Column 1: Galaxy name, Column 2: Observation ID, Column 3: Atomic fine-structure line, Column 4:
Spectral Resolution of PACS at the emission line wavelength, Column 5: Angular resolution of PACS at the emission
line wavelength, Column 6: Exposure time, Column 7: PACS Astronomical Observation Template, Column 8: Size of
observation chopper throw, Column 9: Program ID.
122
NGC 3079 is well known for its spectacular super-bubbles. In the optical
(e.g. Cecil et al., 2001; Veilleux et al., 1994), only the eastern bubble of NGC 3079?s
double-lobed radio structure (Irwin et al., 2019) is visible. The galaxy disk is inclined
such that the E side is nearest to us (Filippenko & Sargent, 1992; Yamauchi et al.,
2004) and the counter bubble is located behind the disk where optical line emission
is extinguished. The ionized outflow originating from the nucleus reaches velocities
of up to ? 1500 km s?1 (Cecil et al., 2001). H? + [N II] emission is observed in
alignment with the bridges and loops seen in radio (Duric et al., 1983; Filippenko &
Sargent, 1992). There is a slow-moving (v ? 0.1c) parsec-scale radio jet (Middelberg
et al., 2005) and copious amounts of hot X-ray gas that are aligned with the optical
emission-line filaments (Cecil et al., 2002).
NGC 4945 harbors a radio-quiet AGN and a nearly edge-on circumnuclear
starburst disk with ? 83 pc radius (Marconi et al., 2000a). It is known for its high
obscuration along the line of sight (NH ? 3.5?1024 cm?2; Puccetti et al., 2014) and
is the brightest Seyfert 2 galaxy in the hard X-ray range (> 20 keV; Itoh et al., 2008),
radiating at a variable rate of L/LEdd ? 0.1 (Madejski et al. 2000). H? emission
tracing a central outflow cone was detected by Lehnert (1992); the cone has been
imaged in both optical and near-infrared emission lines by Moorwood et al. (1996).
In the soft X-ray (? 0.6 keV), the plume is interpreted as a mass-loaded wind
launched from the nuclear starburst (Schurch et al., 2002). Heckman et al. (1990)
detected line splitting in optical spectroscopic data and found the line emission to
originate from the surface of an expanding conical structure. Line splitting starts
around 70 pc and extends out to 700 pc.
123
[OI] 63 [OIII] 88 [NII] 122 [OI] 145 [CII] 158
45
69? 41? 00?? 37
28
45??
20
69? 40? 30?? 12
4
56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s
34
?43? 00? 45??
27
?43? 01? 00??
20
?15?? 14
7
?30??
0
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
41
?65? 20? 00??
33
?15?? 25
16
?30??
8
?45??
0
12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s
37
?25? 17? 00??
29
? 2215??
15
?30??
7
0
34s 33s 0h 47m 32s 34s 33s 0h 47m 32s 34s 33s 0h 47m 32s 34s 33s 0h 47m 32s 34s 33s 0h 47m 32s
32
?0? 00? 30??
25
19
?45??
13
?0? 01? 00?? 6
0
42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s
62
49
37
25
13
0
30s 28s 13h 05m 26s
55
?49? 27? 45??
44
?49? 28? 00?? 33
21
?15??
10
0
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.1 Signal-to-noise ratio maps. Contours are Contours are 0.1, 0.3, 0.5, 0.7,
0.8, and 0.9 of the peak value in each image. Black crosses mark the adopted galaxy
center.
124
NGC 4945 NGC 3079 NGC 1068 NGC 253 Circinus Cen A M 82
[OI] 63 [OIII] 88 [NII] 122 [OI] 145 [CII] 158
26.84 15.60 3.48 2.06 16.69
69? 41? 00?? 21.79 12.65 2.83 1.67 13.62
16.73 9.69 2.19 1.27 10.55
45??
11.68 6.74 1.55 0.88 7.48
69? 40? 30?? 6.62 3.78 0.90 0.48 4.42
175 pc
1.57 0.83 0.26 0.09 1.35
56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s
3.62 0.37 0.15 0.25 2.16
2.93 0.31 0.12 0.20 1.80
?43? 01? 00??
2.24 0.25 0.10 0.15 1.43
?15?? 1.55 0.18 0.07 0.11 1.06
0.87 0.12 0.05 0.06 0.70
?30?? 184 pc
0.18 0.06 0.03 0.01 0.33
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
6.90 1.70 0.60 0.53 4.30
?65? 20? 00??
5.57 1.38 0.49 0.43 3.48
?15??
4.24 1.06 0.37 0.32 2.66
? ?? 2.90 0.74 0.26 0.22 1.8530
1.57 0.42 0.15 0.12 1.03
?45?? 204 pc
0.24 0.10 0.03 0.01 0.21
12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s
11.58 2.70 2.77 1.45 8.97
?25? 17? 00??
9.31 2.17 2.23 1.16 7.26
?15?? 7.04 1.64 1.70 0.88 5.56
4.77 1.11 1.16 0.59 3.85
?30??
2.50 0.59 0.62 0.31 2.14
189 pc
0.24 0.06 0.08 0.02 0.43
34s 33s 0h 47m 32s 34s 33s 0h 47m 32s 34s 33s 0h 47m 32s 34s 33s 0h 47m 32s 34s 33s 0h 47m 32s
3.30 2.08 0.53 0.20 1.77
?0? 00? 30??
2.69 1.69 0.44 0.16 1.49
?45?? 2.08 1.29 0.35 0.12 1.21
1.47 0.90 0.25 0.08 0.93
?0? 01? 00?? 0.85 0.51 0.16 0.05 0.65
698 pc
0.24 0.11 0.07 0.01 0.37
42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s
1.81
55? 41? 00?? 1.45
1.09
45??
0.73
55? 40? 30??
0.37
776 pc
0.01
10h 02m 00s 58s 10h 01m 56s
?49? 27? ?? 4.76 1.71 1.97 1.65 7.6845
3.85 1.38 1.59 1.32 6.20
?49? 28? 00??
2.93 1.04 1.20 0.99 4.72
2.02 0.71 0.81 0.67 3.25
?15??
1.10 0.38 0.43 0.34 1.77
184 pc
0.19 0.04 0.04 0.01 0.29
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.2 Total integrated emission line fluxes in 10?17 W m?2. Contours are 0.1,
0.3, 0.5, 0.7, and 0.9 of the peak flux in the image. Black crosses mark the adopted
galaxy center.
125
NGC 4945 NGC 3079 NGC 1068 NGC 253 Circinus Cen A M 82
4.4 Observations, Archival Data, and Spectral Analysis
The observations of the fine-structure lines, [O I] 63, [O III] 88, [N II] 122,
[O I] 145, and [C II] 158 were performed with the Photoconductor Array Camera
and Spectrometer (PACS; Poglitsch et al., 2010) on board the ESA Herschel Space
Observatory (Pilbratt et al., 2010). These observations were done in pointed ob-
serving mode with PACS line and range scan spectroscopy astronomical observ-
ing templates (AOTs). A large chopper throw is used for a majority of the ob-
servations (two medium throws were used for Cen A ObsID 134220258 and NGC
3079 ObsID 134221391). Observations of M 82 are from the Science Demonstration
Phase SDP esturm 3 (PI: E. Sturm). Observations of Cen A, Circinus, NGC 253,
and NGC 4945 are from Guaranteed Time Key Programs KPGT esturm 1 (PI: E.
Sturm) and KPGT rguesten 1 (PI: R. Gu?sten). NGC 3079 is from the Director?s
Discretionary Time DDT estrum 4 (PI: E. Sturm). More specific observation details
are listed in Table 4.3.
The FIR data were retrieved from the Herschel Science Archive (HSA)1 via
the Herschel Interactive Processing Environment (HIPE v15.0.0; Ott, 2010). These
data have been pipeline-processed at the Herschel Science Centre with the Standard
Product Generation (SPG) software v14.2.0 up to the rebinnedCube task. The
standard reduction steps include glitch masking, bad and noisy pixel masking, dark
and background subtraction, spectral flatfielding, and flux calibration to Jy per
spaxel. The uncertainty on the absolute flux calibration is of the order ? 6% and
1http://www.archives.esac.esa.int/hsa/whsa/
126
the wavelength calibration uncertainties for the red and blue channel are ? 20 km
s?1 and ? 40 km s?1, respectively (Poglitsch et al., 2010).
For each spaxel in an observation, a second-order polynomial is fit to the
continuum and subtracted from the spectrum. Two Gaussian components were then
fit to the continuum-subtracted spectrum and line profile properties are estimated
using the sum of the two components. It should be noted that NGC 1068 needed
three Gaussian components to more accurately capture the broad wings exhibited
in its emission lines, and in this case, the sum of the three Gaussian components is
used to estimate line profile properties. Signal-to-noise ratio maps are presented in
Table 4.3, where in general, the largest SNRs are observed in [C II] 158 and [O I] 63 .
The velocity-integrated emission line flux and ratio maps are discussed in Section 4.5
and the kinematic analysis of these data is presented in Section 4.6.
We note that the observed wavelength range of [N II] 122 contains a second
pass ghost in most of the non-central PACS spectrometer spaxels. This ghost orig-
inates from the [C II] 158 line and appears as a broad spectral feature shifted red-
ward of the [N II] 122 line profile. This ghost lies close to the [N II] 122 line profile
and leaves a small amount of observable continuum red-ward of [N II] 122 making
continuum placement difficult. Therefore, we exclude the analysis of [N II] 122 in
our discussion of the outflows in Section 4.9.
127
4.5 Velocity-Integrated Emission Line Flux and Ratio Maps
Figure 4.2 and Figure 4.3 show the maps of the velocity-integrated line fluxes
and line ratios, respectively. These emission line ratios are useful diagnostic tools
that can probe the physical properties of the ISM. The spatial distribution of emis-
sion ratios can reveal the excitation and ionization conditions in different regions
of a galaxy. Moreover, it is possible to infer the ionization or excitation source of
the gas. Line ratios constrain physical gas properties such as the strength of the
surrounding radiation field, chemical abundances, local temperatures, and gas den-
sities. To compute these line ratio maps, the total integrated flux maps were first
smoothed to the same resolution and aligned to a common pixel grid. Here, we first
briefly summarize the physical conditions traced by the following emission line ratios:
[O I] 63/[C II] 158, [O III] 88/[O I] 63, [O III] 88/[N II] 122, [O III] 88/[C II] 158,
[O I] 145/[O I] 63, and [C II] 158/[N II] 122, before discussing the results.
It should be noted that [O I] 63 can be affected by self-absorption from cold
foreground gas with column densities as low as NH ? 2? 1020 cm?2 (Liseau et al.,
2006). Therefore, any conclusions drawn from ratio values dependent upon the
[O I] 63 emission line flux should be made with care. In the ratios listed below, if self-
absorption is present in the [O I] 63 spectrum, then the [O I] 63 / [C II] 158 values
will be underestimated while the [O III] 88 / [O I] 63 and [O I] 145 / [O I] 63 values
will be overestimated. In this galaxy sample, [O I] 63 self-absorption is only ob-
vious in the spectra along the NE portion of the galaxy major axis of NGC 4945
(absorption is not obviously detected above and below the disk).
128
4.5.1 Constraints from the Emission Line Ratios
[O I] 63 / [C II] 158 ? [C II] 158 and [O I] 63 are the dominant coolants in
PDRs, so the relative strength of their emission is a proxy for the FUV heating
intensity (a measure of how many FUV photons contribute to the gas heating) in
these regions (Parkin et al., 2014). This ratio can also be used to discriminate
between XDRs near AGN and classical PDRs. PDR emission is produced at the
outer edges of molecular clouds, but at high densities (n ? 105 cm?3) in XDRs, the
deeply penetrating X-ray photons can ionize [O I] in a larger volume of the molecular
cloud, thereby enhancing the [O I] 63 emission (Meijerink et al., 2007). In XDRs,
the [O I] 63 line emission can substantially exceed the emission of [C II] 158 and
become the dominant cooling mechanism in the neutral gas (Herrera-Camus et al.,
2018b). Thus, the ionization effects of the AGN can be traced by high ratios of
[O I] 63 / [C II] 158 .
[O III] 88 / [O I] 63 ? This ratio yields the ionized to neutral gas ratio in a
region, but the ratio alone does not distinguish whether the source of ionization is the
AGN or SB (Ferna?ndez-Ontiveros et al., 2016). Low ratios indicate the dominance
of the [O I] 63 line emission from dense, neutral gas.
[O III] 88 / [N II] 122 ? [O III] 88 and [N II] 122 line emission can arise in
the NLR of AGN. Thus, their line ratio is a sensitive indicator of the strength of the
ionization parameter, U , (the number of incident ionizing photons divided by the
hydrogen gas density within the NLR of an AGN; Abel et al., 2009) of the NLR.
In stellar H II regions, this ratio directly measures the effective temperature of the
129
stars responsible for ionization. The ratio reveals the stellar classification of the
youngest stars in the region (Ferkinhoff et al., 2011).
[O III] 88 / [C II] 158 ? This ratio is an excitation sensitive indicator.
[O III] 88 line emission arises in the NLR of AGN or in H II regions where hot O
and B stars produce photons energetic enough to ionize [O III] 88 . Therefore it
traces regions of high radiation fields and gives a measure of the relative ionized to
neutral gas abundances.
[O I] 145 / [O I] 63 ? In the optically thin limit ([O I] 145 / [O I] 63 values
. 0.1; Tielens & Hollenbach, 1985), and in the temperature range 100 ? 400 K
and density range nH. 104 cm?3, the [O I] 145 / [O I] 63 line ratio is an effec-
tive temperature tracer for the neutral gas in PDRs (Liseau et al., 2006; Tielens
& Hollenbach, 1985). Higher values of this ratio indicates high gas temperatures
(Ferna?ndez-Ontiveros et al., 2016; Kaufman et al., 1999). The gas temperature
traced by this ratio is anti-correlated with the gas density traced by [S III] 33 ?m/
[S III] 18 ?m (Spinoglio et al., 2015).
[C II] 158 / [N II] 122 ? [N II] 122 line emission arises solely from low-
excitation photoionized gas, so the [C II] 158 / [N II] 122 ratio can quantify the
fraction of [C II] 158 emission originating from the neutral gas in PDRs.
4.5.2 Results from the Emission Line Ratios
M 82: Emission in all lines (see Figure 4.2) show elongation from E to W,
however, [O I] 63 and [O III] 88 show a slightly more compact nuclear morphology.
130
[OI]63/[CII] [OIII]/[OI]63 [OIII]/[NII] [OIII]/[CII] [OI]145/[OI]63 [CII]/[NII]
2.03 0.71 4.91 1.23 0.13 8.31
69? 41? 00??
1.70 0.61 4.37 1.02 0.11 7.02
45?? 1.37 0.50 3.84 0.81 0.08 5.73
? ? ?? 1.04 0.40 3.31 0.60 0.06 4.4469 40 30
0.71 0.30 2.78 0.39 0.04 3.15
56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s
1.72 0.64 3.47 0.29 0.10 16.72
?43? 01? 00?? 1.38 0.50 3.04 0.25 0.09 14.75
?? 1.03 0.37 2.61 0.21 0.08 12.77?15
0.69 0.23 2.18 0.18 0.07 10.80
?30??
0.35 0.10 1.75 0.14 0.05 8.82
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
1.72 0.93 4.77 1.18 0.09 9.65
?65? 20? 00??
1.42 0.75 3.91 0.93 0.08 8.15
?15??
1.11 0.58 3.04 0.68 0.07 6.65
?30??
0.81 0.40 2.17 0.43 0.06 5.14
?45??
0.50 0.22 1.31 0.18 0.04 3.64
12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s 12s 09s 14h 13m 06s
?25? 17? 00??
1.41 0.30 1.53 0.34 0.17 7.53
1.18 0.25 1.29 0.28 0.14 6.30
?15??
0.94 0.21 1.04 0.22 0.11 5.07
?30?? 0.70 0.16 0.80 0.16 0.08 3.84
0.46 0.12 0.56 0.10 0.06 2.62
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
?0? 00? 30?? 1.96 0.95 4.22 1.21 0.08 7.01
?40?? 1.56 0.78 3.37 0.94 0.07 6.02
? ?? 1.15 0.62 2.53 0.68 0.06 5.0450
? ? ? ?? 0.75 0.46 1.68 0.42 0.05 4.050 01 00
0.34 0.30 0.83 0.15 0.04 3.07
42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s
?49? 27? 45?? 1.32 0.38 1.30 0.44 0.35 10.40
1.07 0.32 1.04 0.35 0.27 8.27
?49? 28? 00??
0.82 0.25 0.79 0.25 0.19 6.14
?15??
0.57 0.19 0.54 0.16 0.11 4.01
0.32 0.13 0.29 0.06 0.03 1.88
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.3 Maps of the emission line ratios. Contours are 0.1, 0.3, 0.5, 0.7, 0.8, and
0.9 of the peak value in each image.
This morphology is emphasized in Figure 4.3 where high ratios of [O I] 63 /[C II] 158 ,
[O III] 88 /[N II] 122 , and [O III] 88 /[C II] 158 are in a compact region around the
galaxy nucleus. The emission line ratios also indicate that the neutral gas traced by
[O I] 63 and the ionized gas traced by [O III] 88 and [N II] 122 is typically stronger
near the galaxy center, but further from the center, [O I] 145 and [C II] 158 emission
begin to dominate.
C13 shows that the bipolar outflow is best determined from the
[O III] 88 /[O I] 63 ratio. Our smaller FOV makes a direct morphological compar-
131
NGC 4945 NGC 1068 NGC 253 Circinus Cen A M 82
[OI] 63 [OIII] 88 [NII] 122 [OI] 145 [CII] 158
? ? ? ?? 13543 00 40
93
?43? 01? 00??
51
? 920??
-32
? ?? 184 pc40
-75
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
169
?65? 20? 00?? 112
56
?20??
?40??
-57
204 pc
-113
12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s
?? 18115
131
69? 41? 00??
81
45??
31
30??
-18
175 pc
69? 40? 15??
-68
55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s
166
?25? 17? 00?? 115
?15?? 64
13
?30??
-37
? ?? 189 pc45
-88
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
201
?0? 00? 30?? 137
? 7345??
10
?0? 01? 00??
-53
698 pc
?15?? -116
42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s
168
15??
109
55? 41? 00??
51
45??
-7
30??
-65
776 pc
55? 40? 15?? -124
10h 02m 00s 10h 01m 57s
150
?49? 27? 45??
96
?49? 28? 00?? 41
? -1315??
-68
?30?? 184 pc
-123
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.4 Maps of the median velocities, v50 in units of km s
?1 . Contours are in
eight equal steps between the minimum velocity and the maximum velocity in each
image.
132
NGC 4945 NGC 3079 NGC 1068 NGC 253 M 82 Circinus Cen A
[OI] 63 [OIII] 88 [NII] 122 [OI] 145 [CII] 158
? 29843? 00? 40??
238
?43? 01? 00??
178
?20?? 119
59
? ?? 184 pc40
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
289
?65? 20? 00?? 231
173
?20??
115
?40??
57
204 pc
12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s
?? 27315
219
69? 41? 00??
164
45??
109
30??
54
175 pc
69? 40? 15??
55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s
429
?25? 17? 00?? 362
?15?? 294
226
?30??
159
? ?? 189 pc45
91
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
469
?0? 00? 30?? 375
?45?? 281
187
?0? 01? 00??
93
698 pc
?15??
42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s
385
15??
308
55? 41? 00??
231
45??
154
30??
77
776 pc
55? 40? 15??
10h 02m 00s 10h 01m 57s
520
?49? 27? 45??
430
?49? 28? 00?? 341
? ?? 25215
163
?30?? 184 pc
73
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.5 Maps of the 1 ? ? line widths, W ?11? , in units of km s . Contours are
0.4, 0.5, 0.6, 0.7, 0.8, and 0.9 of the peak width in each image.
133
NGC 4945 NGC 3079 NGC 1068 NGC 253 M 82 Circinus Cen A
ison difficult because our FOV encapsulates mostly the base of the outflow cones.
C13 reports that the highest [O III] 88 / [O I] 63 ratios are highest in the N portion
of the SB region, which is consistent with our findings. C13 also reports an asym-
metry in the disk in the [O I] 145 / [O I] 63 line ratios. Our results are consistent
with their findings with the higher ratios relegated to the E portion of the disk and
the smaller ratios in the W portion of the disk.
Cen A: In Figure 4.2 the adopted galaxy center of Cen A is spatially co-
incident with the peak emission in all lines. Cen A exhibits a centrally condensed
nuclear emission of neutral gas with a slightly shallower flux gradient towards the
NE of the nucleus, as seen in the [O I] 63 and [O I] 145 maps. On the other hand,
the ionized gas traced by [O III] 88 and [N II] 122 shows bright, elongated nuclear
emission (oriented SE to NW) surrounded by diffuse emission that extends to the
edges of the PACS FOV. Although [C II] 158 is seen as diffuse emission outside
of the galaxy center similar to [O III] 88 and [N II] 122 , the brightest contours of
[C II] 158 are not elongated SE to NW as seen in the ionized gas. Instead, the
brightest contours of [C II] 158 emission are closer in morphology to the compact
nuclei seen in the neutral gas.
NE of the galaxy nucleus, in a location spatially coincident with ?Knot A?,
there is a compact region of high [O III] 88 /[C II] 158 ratios. The highest
[O I] 63 /[C II] 158 ratios in Figure 4.3 for Cen A are localized to the galaxy center,
but there is some slight extension of higher ratios to the NE in the direction of the jet.
There is also elongation of higher [C II] 158 /[N II] 122 ratios to the NE in the direc-
tion of the jet. In general, the neutral gas traced by [O I] 63 dominates the central re-
134
gions (as seen in the low [O III] 88 /[O I] 63 ratios and the high [O I] 63 /[C II] 158 ratios).
Circinus: The nuclear neutral gas emission seen in [O I] 63 and [O I] 145 shows
a compact spatial distribution with a shallower flux gradient extending to the SW
of the galaxy center. The ionized gas traced by [O III] 88 in contrast shows elon-
gated emission with a shallower flux gradient extending to the NW. We see that the
emission in [N II] 122 and [C II] 158 are the most spatially extended of all the lines
and show a preferential elongation along the north ? south axis.
The well-known ionization cone in Circinus is most clearly identified in the
[O III] 88 /[N II] 122 , [O III] 88 /[C II] 158 , and [C II] 158 /[N II] 122 emission line
ratio maps shown in Figure 4.3. The [O III] 88 /[N II] 122 ratios are, on average, a
factor of 2 greater inside the cone than those in the disk. The high [O I] 63 /[C II] 158
ratios localized to the central region of Circinus reveals the effects of the AGN on
the nearby gas ([O I] 63 line emission can be particularly strong in XDRs or NLRs
near AGN).
NGC 253: The emissions for all of the fine-structure lines in NGC 253
lie in a elongated morphology along the NE ? SW axis, coincident with the 2.2
?mm stellar bar detected by Scoville et al. (1985). There is a steep flux gradient
for all lines towards the SW. Most of the [O I] 63, [O III] 88, and [O I] 145, emis-
sion is concentrated within the central 10???10??. In contrast, the [C II] 158 and
[N II] 122 emission distributions extend further above and below the galaxy major
axis.
Again we see the effects of the AGN radiation on the gas in the elevated
values of the line ratios of [O I] 63 /[C II] 158 and [O III] 88 /[C II] 158 (although
135
these ratios also show some elongation towards the E along the galaxy disk) and the
low ratios of [C II] 158 /[N II] 122 found in compact nuclear regions in NGC 253.
The [O I] 63 /[C II] 158 and [O III] 88 /[C II] 158 ratios also indicate the presence
of neutral gas (as traced by [C II] 158 ) in the N outflow cone of NGC 253.
NGC 1068: All line emission in NGC 1068 is elongated along the NE ?
SW axis and aligned with the galaxy?s 2.5 kpc near-infrared nuclear bar (Schinnerer
et al., 2000). The peak line emission for [O I] 63, [O III] 88, and [O I] 145 are local-
ized to a compact region near the galaxy nucleus while [C II] 158 and [N II] 122 emission
is much more diffuse.
The outflow in NGC 1068 is most obvious in the [O III] 88 /[O I] 63 ratio (see
Figure 4.3) where the largest ratios are in a compact region NE of the galaxy nu-
cleus in the direction of the known ionization cone and outflow. The effects of
the AGN radiation are seen in the nuclear compact morphologies of high ratios in
[O I] 63 /[C II] 158 , [O III] 88 /[N II] 122 , and [O III] 88 /[C II] 158 . Also clear
from those ratios is that the neutral gas traced by [C II] 158 increases in emission
further away from the galaxy center. The higher ratios in [O I] 145 /[O I] 63 N of
the nucleus and the higher ratios of [C II] 158 /[N II] 122 S of the nucleus appear to
trace molecular arms in NGC 1068.
NGC 3079: [C II] 158 line emission in NGC 3079 is diffuse and elongated
in the direction of the galaxy major axis. East of the nucleus, a steep flux gradient
is observed along the galaxy minor axis in the direction of the optical bubble (Cecil
et al., 2001; Veilleux et al., 1994).
NGC 4945: For all lines in NGC 4945, peak line emission is in a compact
136
region near the nucleus. The [O III] 88, [N II] 122, and [C II] 158 line emission is
elongated N to S where [N II] 122 and [C II] 158 are more diffuse than [O III] 88 .
The neutral gas traced by [O I] 63 is diffuse but the flux decreases more uniformly
with increasing distance from the nucleus, whereas the steeper flux gradients of
[O III] 88, [N II] 122, and [C II] 158 lie in the direction of the galaxy minor axis.
The [O I] 145 emission is elongated to the S and the SE of the nucleus. It should be
noted that there is significant self-absorption in [O I] 63 for NGC 4945. We do not
correct for this absorption, so the [O I] 63 fluxes reported here should be considered
lower limits.
The [O III] 88 /[N II] 122 ratios in NGC 4945 (see Figure 4.3) show high val-
ues in a compact region inside the NW outflow cone and low ratios in the SE cone,
indicating a higher ionization in the NW cone. The [O III] 88 /[O I] 63 ratios are
highest in the disk along the NE direction. As mentioned earlier in this section, the
[O I] 63 spectrum of NGC 4945 suffers from self-absorption along the NE portion
of the disk. So the higher [O III] 88 /[O I] 63 values are most likely due to the self-
absorption as opposed to higher ionization. The [O I] 145 /[O I] 63 ratio shows high
values in a compact region around the nucleus which also extends SE in the direction
of the outflow. Lower [O I] 145 /[O I] 63 ratios in the NW cone suggest a lower tem-
perature for the neutral gas compared to the SE cone. The [O I] 63 self-absorption
is not obviously observed in regions above and below the galaxy disk. Therefore,
even though the [O I] 145 /[O I] 63 values near the nucleus are most likely overesti-
mates, there is greater confidence in the accuracy of the [O I] 145 /[O I] 63 values in
the NW and SE outflow cones of NGC 4945. The [O III] 88 /[C II] 158 ratios show
137
low values in a compact region around the galaxy center indicating higher ioniza-
tion near the AGN with [C II] 158 line emission increasing with distance from the
nucleus. The [C II] 158 /[N II] 122 ratios are lower in a compact region around the
nucleus compared to a compact region in the NW outflow cone with higher values,
indicating a higher [C II] 158 emission in the NW outflow.
4.6 Gas Kinematics
The widths and velocities of the emission line profiles in each spaxel are de-
scribed by the non-parametric characteristics v50 and W1? . v50 is the median velocity
of the fitted emission line profile, i.e. 50% of the emission is produced at velocities
below v50 . Zero velocity corresponds to the rest wavelength at the galaxy?s systemic
velocity. W1? is the width of the line profile within 1-? standard deviation of v50 .
Figure 4.4 and Figure 4.5 show the v50 and W1? maps of the sample galaxies,
respectively. These maps are analyzed in Section 4.8 and the results of this analysis
are discussed in Section 4.9 where they are compared to the multi-phase outflows
known to exist in these galaxies.
4.7 PDR Modeling
It is possible that the emission of the neutral ISM component in the outflows of
our galaxies originate from photodissociation regions (PDRs). PDRs are the warm,
dense surfaces of molecular clouds that are exposed to the far-ultraviolet (FUV)
photons (6 eV < h? < 13.6, Hollenbach & Tielens, 1997, 1999) that escape from H II
138
Solution A Solution B Solution C Solution D
?43? 00? 45?? 1780 1780 56200 56200
1426 1426 48505 48505
?43? 01? 00??
1072 1072 40811 40811
?15??
718 718 33116 33116
?30??
364 364 25422 25422
med: 17 mean: 188 med: 562 mean: 709 med: 17800 mean: 20636 med: 31600 mean: 31497
10 10 17727 17727
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
1225 1225 100000 100000
?65? 20? 00??
982 982 86213 86213
?15?? 739 739 72427 72427
? ?? 496 496 58641 5864130
253 253 44855 44855
?45??
med: 178 mean: 179 med: 562 mean: 630 med: 31600 mean: 36449 med: 31600 mean: 49763
10 10 31069 31069
12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s
15?? 3160 3160 100000 100000
2530 2530 86200 86200
69? 41? 00??
1900 1900 72400 72400
45??
1270 1270 58600 58600
69? 40? 30??
640 640 44801 44801
med: 562 mean: 581 med: 1780 mean: 1476 med: 31600 mean: 39309 med: 56200 mean: 66155
10 10 31001 31001
56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s
1065 1065 178000 178000
?25? 17? 00?? 854 854 145960 145960
?? 643 643 113920 113920?15
432 432 81880 81880
?30??
221 221 49840 49840
?45?? med: 100 mean: 151 med: 562 mean: 663 med: 56200 mean: 53769 med: 56200 mean: 66020
10 10 17800 17800
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
3160 3160 178000 178000
?0? 00? 30?? 2530 2530 145960 145960
1900 1900 113920 113920
?45??
1270 1270 81880 81880
?0? 01? 00??
640 640 49840 49840
?15?? med: 56 mean: 117 med: 316 mean: 831 med: 31600 mean: 35731 med: 31600 mean: 45638
10 10 17800 17800
42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s
3160 562000
?49? 27? 45??
2530 451600
?49? 28? 00?? 1900 341200
?15??
1270 230800
640 120400
?30?? med: 562 mean: 774 med: 178000 mean: 227626
10 10000
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.6 Maps of the hydrogen density, nH in each object. Units are in cm
?3.
The ?low? density range is defined as 0 < n < 104 cm?3H and the ?high? density
range is defined as 104 < nH< 10
7 cm?3. See Section 4.7.4 for details about the
determination of these density ranges and definitions of Solutions A?D. Solution A
is the low density solution with only the [C II] 158 flux correction (see Section 4.7.1
and Section 4.7.4), Solution B is the low density solution with both the fluxes of
[C II] 158 and [O I] 63 corrected (see Section 4.7.1, Section 4.7.2 and Section 4.7.4),
Solution C is the high density solution with only the [C II] 158 flux correction,
and Solution D is the high density solution with both the fluxes of [C II] 158 and
[O I] 63 corrected. As discussed in Section 4.7.4, the high density PDR solutions
?C? and ?D? lead to unphysical ISRFs. Therefore, we have excluded them from the
analysis of the outflow.
139
NNGGCC 44994455 NNNNGGGGCCCC 1111000066668888 NNNNGGGGCCCC 222255553333 MMMM 88882222 CCCCiiiirrrrcccciiiinnnnuuuussss CCCCeeeennnn AAAA
Solution A Solution B Solution C Solution D
?43? 00? 45?? 1780 1780 1.79 1.79
1426 1426 1.63 1.63
?43? 01? 00??
1072 1072 1.47 1.47
?15??
718 718 1.32 1.32
?30??
364 364 1.16 1.16
med: 56 mean: 301 med: 316 mean: 390 med: 1.00 mean: 1.00 med: 1.78 mean: 1.76
10 10 1.00 1.00
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
5620 5620 1.83 1.83
?65? 20? 00??
4496 4496 1.67 1.67
?15?? 3372 3372 1.50 1.50
?30?? 2248 2248 1.33 1.33
1124 1124 1.17 1.17
?45??
med: 562 mean: 1058 med: 1000 mean: 1165 med: 1.00 mean: 1.00 med: 1.78 mean: 1.39
0 0 1.00 1.00
12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s
15?? 5620 5620 1.83 1.83
4496 4496 1.66 1.66
69? 41? 00??
3372 3372 1.50 1.50
45??
2248 2248 1.33 1.33
69? 40? 30??
1124 1124 1.17 1.17
med: 1780 mean: 2186 med: 1780 mean: 1902 med: 1.00 mean: 1.00 med: 1.78 mean: 1.72
0 0 1.00 1.00
56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s
31600 31600 1.78 1.78
?25? 17? 00?? 25286 25286 1.54 1.54
18972 18972 1.29 1.29
?15??
12658 12658 1.05 1.05
?30??
6345 6345 0.81 0.81
?45?? med: 1000 mean: 3389 med: 3160 mean: 4572 med: 1.00 mean: 0.96 med: 1.00 mean: 1.18
31 31 0.56 0.56
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
10000 10000 1.78 1.78
?0? 00? 30?? 8000 8000 1.62 1.62
6000 6000 1.46 1.46
?45??
4000 4000 1.31 1.31
?0? 01? 00??
2000 2000 1.15 1.15
?15?? med: 316 mean: 891 med: 316 mean: 593 med: 1.00 mean: 1.00 med: 1.00 mean: 1.23
0 0 0.99 0.99
42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s 42s 41s 40s 2h 42m 39s
17800 17
?49? 27? 45??
14243 14
?49? 28? 00?? 10687 11
?? 7130 8?15
3574 5
?30?? med: 3160 mean: 4722 med: 5 mean: 7
17 1
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.7 Maps of the strength of the UV radiation field, G0, for the low density
and high density limits PDR solutions. Definitions of the Solutions are the same as
those in Figure 4.6.
140
NNGGCC 44994455 NNNNGGGGCCCC 1111000066668888 NNNNGGGGCCCC 222255553333 MMMM 88882222 CCCCiiiirrrrcccciiiinnnnuuuussss CCCCeeeennnn AAAA
regions in star-forming regions or the accretion disks in AGN. The absorption of FUV
photons by gas and dust is converted to FIR radiation including intense emission
of [C II] 158 and [O I] 63 transitions, mainly excited by collisions with H2 molecules
and H I (nH & 103 ? 104 cm?3;Liseau et al., 2006), whose line emissions dominate
the gas cooling in PDRs (Tielens & Hollenbach, 1985). These two transitions are
the brightest PDR emission lines in the Herschel -PACS spectral range and thus, are
sensitive probes to the physical conditions of the gas in a PDR.
In this section, we use the fluxes of [O I] 63, [O I] 145, and [C II] 158 and the
total infrared continuum emission (TIR ? 30 ? 1000 ?m) to derive simple PDR
models and infer the spatially-averaged properties of the gas by comparing observed
emission line ratios to theoretically predicted line ratios. The models we employ here
were first presented by Hollenbach et al. (1991); Tielens & Hollenbach (1985); Wolfire
et al. (1990). Kaufman et al. (2006, 1999) has since updated these models to include
more recent values of atomic and molecular data, current chemical rate coefficients,
and grain photolelectric heating rates. These models assume that the light-emitting
material is a plane-parallel, semi-infinite slab illuminated from one side. We use
the PDR Toolbox1 (Pound & Wolfire, 2008), a publicly available implementation of
the model diagnostics to constrain the intensity of the ambient UV radiation field
(G0 or ISRF; conventionally expressed in units of the local Galactic FUV flux =
1.6?10?3 erg cm?2 s?1 , a.k.a Habing flux, Habing, 1968) and the hydrogen nucleus
density (nH). Before we can estimate the PDR parameters of our sample, we must
first consider a flux correction for [C II] 158 and [O I] 63 .
1http://dustem.astro.umd.edu/pdrt/
141
4.7.1 [C II] 158 Emission
The [C II] 158 line emission originates from both the ionized and neutral ISM
phases due to the transition?s low ionization potential (11.26 eV), low transition
temperature (91 K), and low critical density for collisions (2 ? 103 cm?3). The
PDR model, however, only considers the [C II] 158 flux arising from the neutral
gas. Therefore, we must correct for the fraction of [C II] 158 flux that is emitted
from the ionized gas. We can derive this fraction using the [N II] 122 flux whose
higher ionization potential (14.53 eV) implies that the nitrogen emission originates
solely from the ionized gas. We assume that the [C II] 158 flux from the ionized gas
scales directly with that of the [N II] 122 line using the relation valid for dense HII
regions and for a Milky Way C/N abundance ratio equal to 3.8 (Rubin et al., 1993,
1988):
[CII]ionized = 1.1? [NII]. (4.1)
4.7.2 [O I] 63 Emission
The PDR models employed here assume that the line emission originating
from the irradiated side of a cloud is optically thin. However, when a spectral line
is optically thick, as is the case with [O I] 63 , and the cloud is lit from behind, we
will not observe emission of that line. This means we only see the [O I] 63 flux if
the irradiated side of a cloud is facing us. Kaufman et al. (1999) states that we will
observe half of the [O I] 63 emission produced with the other half radiating away
from our line of sight. We therefore correct for this geometrical effect by multiplying
142
the observed [O I] 63 flux by a factor of two. Most likely, the true [O I] 63 flux
correction factor lies somewhere between one and two. Therefore, we estimate the
PDR solutions with and without the [O I] 63 flux correction. The final PDR results
represent the range of values that the observed gas may have.
As mentioned in Section 4.5, [O I] 63 can suffer from self-absorption which
would also affect the amount of the emission detected. There is heavy absorption
in [O I] 63 in NGC 4945 (see Section 4.5.2), therefore we do not compute a PDR
solution for NGC 4945 using the [O I] 63 flux. The other objects do not show
obvious absorption in [O I] 63 , therefore we apply only a geometrical correction
to the [O I] 63 emission.
4.7.3 Total Infrared Emission
We estimate the TIR emission (8? 1000?m) using the specific flux values at
[O I] 63 , [O III] 88 , [O I] 145 , and [C II] 158 from the continua fitted in Section 4.4.
Assuming optically thin emission and a fixed grain emissivity (? = 1), we
fit the continuum fluxes with a modified blackbody and extract the calculated flux
density at 60?m and 100?m. We then compute the FIR emission using the equation
from Helou et al. (1988) :
FIR [Wm?2] =1.2(6? 10
?14
) (4.2)
? 2.58? I60?m[Jy] + I100?m[Jy] .
Using the IR bolometric correction from Dale et al. (2001) we approximate
143
the TIR emission via:
( )
TIR
log = a0 + a1x+ a
2
2x + a3x
3 + a x44 , (4.3)
FIR
where x = log[I60?m/I100?m] and a ? [0.2738, ?0.0282, 0.7281, 0.6208, 0.9118]
assuming a redshift z = 0.
4.7.4 Hydrogen Density and UV Radiation Field
The solutions from the PDR toolbox are degenerate. Therefore, a predicted
emission line ratio may be produced by either a low density environment with a
strong radiation field or a high density environment with a weak radiation field.
For our sample of galaxies, inspection of the ?2 map in the nH?G0 parameter
space shows a delineation between the degenerate solutions of the PDR modeling.
Therefore, we also modeled the PDR regions within a ?low? density range, 0 <
n < 104 cm?3H and a ?high? density range, 104 < nH< 107 cm?3
Here we present the hydrogen density (nH; see Figure 4.6) and far-ultraviolent
interstellar radiation field (ISRF; G0; see Figure 4.5.2) solutions from the PDR
modeling. Hereafter, for brevity and clarity, the four PDR solutions will be re-
ferred to as Solutions A, B, C, and D. Solution A is the low density solution
with only the [C II] 158 flux correction. The [C II] 158 flux correction attempts to
remove the portion of [C II] 158 emission that originates from ionized gas by us-
ing Equation 4.1. Solution B is the low density solution with both the fluxes of
[C II] 158 and [O I] 63 corrected. The [O I] 63 flux is corrected using a geometrical
144
factor of two as described in Section 4.7.2. Solution C is the high density solution
with only the [C II] 158 flux correction. Solution D is the high density solution
with both the fluxes of [C II] 158 and [O I] 63 corrected.
Setting a high density limit (n > 104H cm
?3) results in a ISRF ? 1G0 across
the entire FOV in the galaxy sample (NGC 4945 is the exception with a peak ISRF
of ? 17G0 around the galaxy center). Radiation fields this weak and this extended
in the central regions of these galaxies are not physical since much higher radiation
fields are expected near an AGN or a circumnuclear starburst. Although NGC
4945 does show a higher ISRF (. 17G0) around the nucleus compared to the other
galaxies, it is still too small to be realistic. Because the radiation fields modeled at
a high density limit produce unrealistic ISRFs, we will only discuss the low-density
solutions A and B in the following analysis.
4.7.5 Caveats
We have derived the spatially-averaged neutral gas properties of our galaxy
sample via a simple and uniform treatment of the integrated line emission fluxes.
However, the physical parameters inferred from any PDR model should not be
taken too literally, since they are dependent upon the assumptions adopted and the
microphysics implemented in a model that oversimplifies a much more complex set
of physical conditions. Moreover, due to the discretization of the parameter space
in the PDR models, we are wary of the pixel-to-pixel variations in the final (nH, G0)
solutions. In particular, if the input flux uncertainties had been lower, it is possible
145
that the final (nH, G0) results would have been different than the results presented
here. This is seen in the final PDR solution maps as a hot/cold pixel in the middle of
an otherwise well defined morphological structure. This is corrected by interpolating
these pixels with surrounding neighbors.
4.8 Methods to Derive the Outflow Properties
In this section the methods for estimating the physical and kinematic outflow
properties in the galaxy sample are outlined. The spectral observation cubes are
first used to model the disk rotation in each galaxy. The resulting velocity field is
then subtracted from the observed emission line profile properties and the residuals
are used to define the spatial location of the outflow. The emission line fluxes within
the wind region are used to estimate line luminosities. The line luminosities in turn,
are used to compute the masses and kinetic energies in the neutral and ionized gas
in the outflow.
4.8.1 Beam Smearing
One of the greatest difficulties in deriving galaxy kinematics from velocity fields
is overcoming the effect of beam smearing (Bosma, 1978). The radial velocities in a
galaxy vary on spatial scales smaller than the beam size of the observing instrument.
Effectively, this means that the velocities at different radii will be blended, thus
flattening the observed velocity field gradients, decreasing the slope of the derived
rotation curve at the galaxy center, and broadening the width of the observed line
146
profiles. This is a significant problem since the broadening effect can be incorrectly
attributed to the gas velocity dispersion.
Fortunately, we can employ methods that utilize the full 3D data cube (two
spatial dimensions and one spectral dimension) of a galaxy and build a model that
directly incorporates the instrument?s 2D point spread function and spectral resolu-
tion. To do this we use the software package 3DBarolo1 (Di Teodoro & Fraternali,
2015) which simulates the observational data in a cube by building a ?tilted-ring?
model that best fits the data. Accounting for the instrumental contribution to the
observed data will result in a disk model which more accurately captures the true
kinematics of the gas. The details of the ?tilted-ring? model and the employment
of 3DBarolo are discussed in the following section.
4.8.2 Tilted Ring Model
The velocity field of a disk galaxy can be described by a ?tilted-ring? model
(Rogstad et al., 1974), where the disk is built from a series of concentric annuli that
increase in radius with the distance from the galaxy center. This model assumes
that the line emitting material is confined to a thin disk and that the kinematics are
dominated by rotational motion. Emission of the gas in each ring is described by
geometric parameters (centroid, radius, width, scale height, inclination angle along
the line of sight of the observer, and position angle of the major axis) and kinematic
parameters (systemic velocity, rotational velocity, and velocity dispersion). For
3DBarolo, the instrumental spectral and spatial resolutions are also inputs so that
1http://editeodoro.github.io/Bbarolo/, version 1.4
147
the final model accounts for both instrumental effects.
In reproducing the observed data cube, 3DBarolo assumes that all of the
velocity dispersion is due to rotation and turbulence. Outflows are known to exist
along the minor axes of the sample galaxies, so this assumption means that the
velocity dispersion of the disk model will be overestimated, with the largest errors at
the center of the galaxy where line broadening suffers the most from beam smearing.
We attempt to mitigate this effect by assuming a constant velocity dispersion
across the disk and estimate the value of this velocity dispersion using data outside
the galaxy center where the beam smearing is the least severe. The data are fit in
two stages. In the first stage, the gas velocity dispersion and the rotational velocity
are left as free parameters. The remaining parameters are input as ?correct? values
and held constant. Afterwards, the mean of the fitted gas velocity dispersions of the
three outmost rings is then computed. For the second stage, this mean dispersion is
input as a fixed parameter and only the rotational velocity is left free. The resultant
disk model is the assumed galaxy velocity field.
4.8.3 Defining the Spatial Location of the Outflow
The kinematics of the outflow can be delineated from those of the disk by
examining the residuals in v50 and in W1? between the observed data and the mod-
eled disk (?v50 and ?W1?, respectively). Recall from Section 4.8.2, the modeled
galaxy velocity field accounts for the instrumental spatial and spectral resolutions.
Therefore, in the residuals, the instrumental effects have been removed. In this
148
galaxy sample, excess line widths are a more reliable indicator of outflow than the
v50 residuals. For example, the outflow of M 82 is known to lie along the galaxy
minor axis. Inspection of the ?W1? maps (see Column 3 in Figure 4.8b) shows an
obvious SE to NW trend consistent with the direction of the outflow, but no such
trend is observed in the ?v50 maps (see Column 3 in Figure 4.8a). For this reason
?W1? is used to define the spatial location of the wind.
The excess line widths are assumed to trace non-circular motions in the gas.
These motions can be due to the presence of morphological features in a galaxy,
such as spiral arms, a nuclear bar, a warped disk, and/or an outflow. Ideally, in
the absence of such features, save the outflow, the spatial location of the wind
would be defined as those regions where ?W ?11? > 0 km s . However, these features
do exist in our galaxy sample and for some, the projected morphologies of those
features overlap with each other, making decomposition of the wind from the features
difficult. Therefore, when defining the spatial location of the wind, it is important to
carefully choose a minimum threshold in ?W1? which would account for gas motions
not related to the outflow.
For various thresholds, features in the ?W1? maps were compared to the
physical features of the galaxies observed at other wavelengths. For example, for
[C II] 158 in NGC 1068, a region of excess line width SW of the nucleus more likely
traces a molecular spiral arm because the region is elongated N to S along the arm
(see Column 3 of Figure 4.19b) and because the S outflow cone lies behind the
galaxy disk where even in the FIR the S cone may be extinguished. A thresh-
old of ?W1? > 25 km s
?1 was chosen to minimize the inclusion of regions where
149
excess line widths were spatially coincident with features unrelated to the known
outflow. Ultimately, this threshold does mitigate the detection of gas motions un-
related to the wind, but it does not remove those motions completely. Although a
higher threshold would more effectively exclude gas motions unrelated to the out-
flow, such a threshold might also exclude the outflow all together. For example,
the detection of a [C II] 158 outflow in M 82 disappears completely with a threshold
of ?W1? > 50 km s
?1 (see Column 5 of Figure 4.8b), however, [C II] 158 is known
to exist in the outflow of M 82 (C13). This is a probable indicator that a thresh-
old of ?W ?11? > 50 km s is too harsh a limit. We include W1? residual maps with
?W > 50 km s?11? as reference for a lower limit to the spatial extent of the wind
in the sample galaxies (see Column 5 of subfigure (b) in Figure 4.8, Figure 4.11,
Figure 4.14, Figure 4.16, Figure 4.19, Figure 4.22, Figure 4.24), but the following
analysis defines the spatial location of the wind as those regions where ?W1? > 25
km s?1.
4.8.4 Line Luminosities
The relation from Solomon et al. (1992) is used to estimate the emission line
luminosities from the observed line fluxes via
Lline
= 1.04? 10?3 Sline ?v (1 + z)?1 ?restD2L, (4.4)L
where Sline ?v is the velocity integrated flux in Jy km s
?1, ?rest = ?obs(1 + z) is the
rest frequency in GHz, and DL is the luminosity distance in Mpc (see Table 4.2 for
150
galaxy details).
4.8.5 Mass of the Neutral Atomic Wind
We follow Hailey-Dunsheath et al. (2010) to estimate the neutral atomic gas
mass in the outflow by using the inferred [C II] 158 luminosity and the PDR solu-
tions (see Section 4.7). Assuming the line emission is optically thin, the minimum
atomic gas mass can be estimated by
( )(
M L ?
)
4
atomic [C II] 158 1.4? 10
= 0.7
M (7 L XC)+ (4.5)
1 + 2e?91K/T? + ncrit/n ,
2e?91K/T
where L[C II] 158 is the [C II] 158 line luminosity in solar units, XC+ is the carbon
abundance per hydrogen atom, n is the gas density, and ncrit is the critical density.
We assume that the PDR surface temperature, TS, derived in Section 4.7 is repre-
sentative of the region?s gas temperature and we adopt XC+ = 1.6 ? 10?4 (Sofia
et al., 2004) which assumes a C+ abundance typical of PDRs in the Milky Way.
The atomic outflow mass estimated here (see Table 4.4) is derived only from
the gas associated with [C II] 158 emission and is not an estimate of the total neutral
atomic gas in the outflow. Also, a lower XC+ value will result in a more massive
wind. Therefore, we emphasize that these mass estimates should be considered lower
limits.
151
4.8.6 Mass of the Ionized Wind
Following Ferkinhoff et al. (2010), the minimum ionized gas mass required to
produce the observed [O III] 88 line emission can be estimated in the high-density
(ncrit/n 1), high-temperature (Teff > 40, 000 K) limit via
M 2ionized gt 4? ?DL mH 1= Fline ? , (4.6)
M gu A h? Xelement Xion
where Fline is the emission line flux, gt = ?giexp (??Ei/kT ) is the partition func-
tion, gu is the statistical weight of the upper transition level, DL is the source
luminosity distance, A is the Einstein A coefficient, ? is the rest frequency of the
emission line, mH is the mass of hydrogen, Xelement is the relative abundance of the
emission line species to hydrogen (e.g. O++/H+, which, for a minimum mass is the
total abundance ratio O/H), and Xion is the fractional abundance of a given species
in a particular ionization state (e.g. O++/O) . We adopt the gas-phase abundance
ratio XO = 5.9? 10?4 from Savage & Sembach (1996).
It is important to note that in our gas mass calculations we are assuming that
all of the oxygen is in the form of O++. The ionization potential of O++ (in this case,
35.12 eV for [O III] 88) is much higher than that of O+ so it is unlikely that all of the
oxygen will be doubly ionized. Therefore, we include an ionization correction term,
Xion, to estimate a more realistic gas-phase abundance ratio, XO++ . For example,
if the stellar T ++eff ? 36, 000 K, the fraction of O to the total oxygen abundance
will only be ? 15%, and the ionized wind mass will be about 7 times larger. The
152
computed masses of the ionized gas in the wind are listed in Table 4.4.
4.8.7 Kinetic Energy of the Wind
The kinetic energy in the wind may be described as the sum of the ?bulk?
kinetic energy and the ?turbulent? kinetic energy of the outflowing gas (Rupke
et al., 2005a; Veilleux et al., 2020) via:
[ ( ) ]2
1 2 W1?,windKEwind = Mwind v50,wind + 3 . (4.7)2 2
The estimates of the kinetic energies in the wind are listed in Table 4.4.
4.9 Properties of the Outflows and Multi-Phase Comparisons
In this section we discuss the results from Section 4.8 and directly compare the
physical properties of the neutral and ionized outflows within each sample galaxy.
For each object, the first figure shows the observed velocity field and 1?? line
width (v50 and W1? , respectively) measured from the line profile fitting outlined in
Section 4.4 and Section 4.6. This figure also shows the modeled v50 and W1? results
from Section 4.8.2, followed by the residuals between the data and the model. Lastly,
the figure shows the regions where the 1-? line width residuals (?W1?) exceed 25 and
50 km s?1 to delineate where the outflow is believed to be significant. Hereafter, the
residual velocities and line widths inside the regions where ?W > 25 km s?11? are
defined as ?v50 wind and ?W1?wind, respectively. The second figure for each object
153
Table 4.4. Wind Properties
Solution A Solution B [O III] 88
Galaxy Name M KE M KE M KE
(1) (2) (3) (4) (5) (6) (7)
Outflow Definition: ?W ?11? > 25 km s
M 82 86.1 34.7 17.4 6.13 3.56 1.62
Cen A 18.2 4.35 1.57 0.27 0.05 0.02
Circinus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 253 106 35.9 10.9 3.57 0.40 0.32
NGC 1068 ? ? ? ? ? ? ? ? ? ? ? ? 8.95 20.8
NGC 4945 ? ? ? ? ? ? ? ? ? ? ? ? 0.36 0.22
Outflow Definition: ?W1? > 50 km s
?1
M 82 ? ? ? ? ? ? ? ? ? ? ? ? 1.22 0.70
Cen A ? ? ? ? ? ? ? ? ? ? ? ? 0.01 0.01
Circinus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
NGC 253 23.6 12.9 1.62 0.93 0.26 0.28
NGC 1068 ? ? ? ? ? ? ? ? ? ? ? ? 5.49 19.0
NGC 4945 ? ? ? ? ? ? ? ? ? ? ? ? 0.07 0.13
Note. ? Masses of the wind are in units of 106 M. Kinetic energies of the
wind are in units of 1053 erg. ?Solution A? and ?Solution B? refer to the low
density limit (nH < 10
4 cm?3) PDR results from Section 4.7.4 used to compute
the neutral outflow mass (see Section 4.8.5) and kinetic energy (see Section 4.8.7).
Outflow masses and kinetic energies listed in the top table are estimated using
fluxes, radial velocities, and 1?? line widths in the spatial regions where the 1??
line width residuals between the observed and modeled velocity field (?W1? ) is
larger than 25 km s?1 (see Section 4.8.3 for further details). The bottom table is
the same as that above except the 1?? line width residual threshold is 50 km s?1.
This higher threshold illustrates the lower limit of the outflow masses and kinetic
energies. Column 1: Galaxy Name. Column 2: Neutral gas mass in the outflow
derived in the low density limit with only the [C II] 158 flux corrected, Column 3:
Kinetic energy of the neutral gas in the outflow derived in the low density limit
with only the [C II] 158 flux corrected, Column 4: Neutral gas mass in outflow
derived in the low density limit with the [C II] 158 and [O I] 63 fluxes corrected,
Column 5: Kinetic energy of the neutral gas in the outflow derived in the low
density limit with the [C II] 158 and [O I] 63 fluxes corrected, Column 6: Ionized
gas mass in the outflow estimated from the [O III] 88 flux (see Section 4.8.6).
Column 7: Kinetic energy of the ionized gas in the outflow estimated from the
[O III] 88 flux (see Section 4.8.7).
154
compares these outflow maps with existing photometric and kinematic data at other
wavelengths. The final figure for each object presents maps of the neutral and ionized
gas masses and the corresponding kinetic energies associated with the outflows.
As mentioned in Section 4.4, the continuum placement of [N II] 122 is difficult
due to the proximity of the second pass ghost to the [N II] 122 line profile. Therefore,
[N II] 122 is excluded from the following analysis because of the high uncertainty in
its line profile measurements. Also, [O I] 63 is not used to model the NGC 4945
galaxy disk due to the presence of strong self-absorption in the line profiles which
is not seen in the other galaxies. Additionally, the wind kinematics for NGC 3079
only include the gas traced by [C II] 158 since this is the only fine structure line
observed in NGC 3079.
4.9.1 M 82
The M 82 v50 maps in all lines (see Figure 4.8a, Column: 1) show that the E
side of the disk is receding from us (largest velocity at ? 145 km s?1 ) and the W
side of the disk is approaching us (largest velocity at ? ?55 km s?1 ). Although the
direction of the disk rotation and the lower approaching radial velocity in this PACS
data is consistent with the analysis in C13, the maximum receding velocity reported
here is ? 45 km s?1 greater than that reported in C13. The M 82 ?v50 wind maps (see
Column 4 of Figure 4.8a) show the NW outflow in all lines and indicate that the cone
is receding from us with ?v50 wind up to ? 65 km s?1 (the largest recession velocity
is in [O I] 145 ). The S outflow is approaching us, but is seen only in the ionized
155
M 82
(a) ?v50 ?v50
v50 Data v50 Model ?v50 (?W1? > 25) (?W1? > 50)
145 145 84 84 84
69? 41? 00??
82 82 46 46 46
45?? 19 19 7 7 7
? ? ?? -43 -43
-30 -30 -30
69 40 30
-107 -107 -68 -68 -68
145 145 84 84 84
69? 41? 00??
82 82 46 46 46
45?? 19 19 7 7 7
-43 -43 -30 -30 -30
69? 40? 30??
-107 -107 -68 -68 -68
145 145 84 84 84
69? 41? 00??
82 82 46 46 46
45?? 19 19 7 7 7
-43 -43 -30 -30 -30
69? 40? 30??
-107 -107 -68 -68 -68
145 145 84 84 84
69? 41? 00??
82 82 46 46 46
45?? 19 19 7 7 7
-43 -43 -30 -30 -30
69? 40? 30?? 175 pc
-107 -107 -68 -68 -68
56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s
(b) W1? Data W1? Model ?W1? ?W1? > 25 ?W1? > 50
223 223 84 84 84
69? 41? 00??
171 171 35 35 35
45?? 119 119 -13 -13 -13
? ? ?? 67 67 -63 -63 -6369 40 30
15 15 -112 -112 -112
223 223 84 84 84
69? 41? 00??
171 171 35 35 35
45?? 119 119 -13 -13 -13
? ? ?? 67 67 -63 -63 -6369 40 30
15 15 -112 -112 -112
223 223 84 84 84
69? 41? 00??
171 171 35 35 35
45?? 119 119 -13 -13 -13
? ? 67 67 -63 -63 -6369 40 30??
15 15 -112 -112 -112
223 223 84 84 84
69? 41? 00??
171 171 35 35 35
45?? 119 119 -13 -13 -13
? ? ?? 67 67 -63 -63 -6369 40 30 175 pc
15 15 -112 -112 -112
56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s 56s 52s 9h 55m 48s
Figure 4.8 M 82: Results from modeling the disk velocity field with
3DBarolo (which accounts for both the instrumental spectral and spatial resolu-
tions) and the location of the outflow based on excess line broadening. Color bar
values are in units of km s?1 . For each line, left to right: observed data, model re-
sult, data - model residual, spatial location of the wind in regions where ?W1? > 25
km s?1 , and spatial location of the wind in regions where ?W1? > 50 km s?1 . Con-
tours in (a) are in five equal steps between the minimum and maximum velocities
in each image. Contours in (b) are 0.3, 0.5, 0.7, and 0.9 of the peak value in each
image. The solid blue line marks the galaxy major axis. The magenta cross marks
the adopted galaxy center.
156
[[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633 [[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633
Figure 4.9 M 82: PACS contours (magenta) overlaid on the total integrated intensity
contours of SiO(2-1) (black, Garc??a-Burillo et al., 2001) and the radio continuum
at 4.8 GHz (grey scale, Wills et al., 1999).
gas traced by [O III] 88 with ?v50 wind velocities ranging from ? ?30 km s?1 to ?
?10 km s?1 . In general, the ?v50 wind velocities increase with increasing (projected)
distance from the galaxy nucleus in [O III] 88 , [O I] 145 , and [C II] 158 . This is
not the case with [O I] 63 , however, where the ?v50 wind velocities are greatest near
the nucleus and decrease further up above the disk.
The largest residual velocity dispersions in the wind region (Column 4 of Fig-
ure 4.8b) are in [O I] 145 with ?W1?wind ranging from ? 25? 85 km s?1 , where the
velocity dispersions increase as the (projected) distance from the nucleus increases.
The range of ?W1?wind velocities is similar between [O I] 63 and [O III] 88 (?W1?wind?
30 ? 65 km s?1 ). Unlike [O I] 145 , the largest residual velocity dispersions in
[O I] 63 are localized in a compact area near the galaxy center and decrease with
increasing distance from the center. [O III] 88 contains a similar compact region near
157
M 82
Mwind KEwind Mwind KEwind
(?W1? > 25) (?W1? > 25) (?W1? > 50) (?W1? > 50)
318.19 160.10
254.76 128.12
69? 41? 00??
191.34 96.14
40?? 127.92 64.16
64.49 32.18
69? 40? 20?? Mtot: 8606.21 M KEtot: 3466.50 ergs Mtot: ? KEtot: ?
1.07 0.20
318.19 160.10
254.76 128.12
69? 41? 00??
191.34 96.14
40?? 127.92 64.16
64.49 32.18
69? 40? 20?? Mtot: 1735.93 M KEtot: 612.59 ergs Mtot: ? KEtot: ?
1.07 0.20
4.99 1.83
4.08 1.48
69? 41? 00??
3.18 1.13
40?? 2.27 0.78
1.36 0.43
69? 40? 20?? Mtot: 930.73 M KEtot: 326.43 ergs Mtot: ? KEtot: ?
0.45 0.08
4.99 1.83
4.08 1.48
69? 41? 00??
3.18 1.13
40?? 2.27 0.78
1.36 0.43
69? 40? 20?? Mtot: 957.11 M KEtot: 336.14 ergs Mtot: ? KEtot: ?
0.45 0.08
1.53 1.53 1.44 1.44
1.24 1.24 1.17 1.17
69? 41? 00??
0.95 0.95 0.90 0.90
40?? 0.66 0.66 0.63 0.63
0.37 0.37 0.36 0.36
69? 40? 20?? Mtot: 356.38 M KEtot: 162.25 ergs Mtot: 122.23 M KEtot: 70.32 ergs
0.08 0.08 0.08 0.08
55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s 55s 9h 55m 50s
Figure 4.10 Mass and KE in the wind derived from the PDR solutions (see Sec-
tion 4.7.4 for definitions of Solutions A, B, C, and D) and the [O III] 88 flux. Units
are in 104M and 1051 erg for the masses and KEs, respectively.
158
[[OOIIIIII]] 8888 SSoolluuttiioonn DD SSoolluuttiioonn CC SSoolluuttiioonn BB SSoolluuttiioonn AA
the nucleus as seen in [O I] 63 , however, although the dispersions in [O III] 88 initially
decrease with increasing distance from the nucleus, at about r ? 10 ??, ?W1?wind increases
at larger radii. [C II] 158 appears to have a near constant ?W1?wind? 40 km s?1 along
the entire length of the collimated structure.
C13 estimate the velocity dispersions are ? 50? 100 km s?1 larger inside the
outflow cones compared to the disk. Our observed data (even in its smaller FOV)
are consistent with this assessment.
The sputtering of dust grains by shocks will release gas-phase silicon. The
silicon is rapidly oxidized and forms SiO (Flower et al., 1996). Therefore, SiO is a
good tracer of shocks, since it is thought to be formed in shocks. Figure 4.9 shows
PACS contours (magenta) overlaid on the SiO (black and white contours) observed
by Garc??a-Burillo et al. (2001) and the 4.8 GHz continuum (greyscale) observed
by Wills et al. (1999). The SiO chimney appears to originate in the galaxy disk
and extends to the NE beyond the PACS FOV. All lines trace this chimney out to
? 20?? (? 350 pc) above the disk and consistently indicate it is receding from us
(see Figure 4.9a - Figure 4.9d). Residuals in [O III] 88 are the most tightly spatially
correlated with the chimney while in [O I] 145 the residuals are the most extended
W and E of the chimney. The residuals in the wind of M 82 are consistent with the
idea that at the base of the outflow, gas flows out of the galaxy disk in filamentary
channels instead of being launched uniformly across the starburst region.
The largest masses and KEs in the neutral outflow lie on the N edge of the
outflow for Solution A, while the largest masses and KEs from Solution B lie inside
the galaxy disk (see Figure 4.10). The morphology of the ionized mass distribution
159
follows that of the [O III] 88 flux distribution (i.e. the overall morphology of the
ionized gas mass in the outflow is ellipsoidal and lies spatially coincident with the
galaxy disk). The highest KEs in the ionized outflow coincide with the portion of
the SiO chimney that intersects the galaxy plane.
C13 estimate the mass of the neutral gas in the outflow to be 2? 8? 107 M;
a value ? 4 ? 15 times lower than the cold molecular gas reported in Walter et al.
(2002, 3.3 ? 108 M) and 3 orders of magnitude higher than the warm molecular
gas reported in Veilleux et al. (2009). C13 estimated a KE of ? 1 ? 5 ? 1054 erg
in the neutral outflow; an order of magnitude lower than that of H? (Shopbell &
Bland-Hawthorn, 1998, 3? 1055 erg) and 2 orders of magnitude larger than the KE
measured in H2.
4.9.2 Cen A
The velocity field shown in the first column of Figure 4.11a indicates that the
E side of the gaseous disk at the center of Cen A is approaching us, while the W
side of the disk is receding from us, consistent with the velocity field observed in the
optical ionized (Bland et al., 1987) and molecular gas (Espada et al., 2009). The
velocities range from v50? ?60 to +105 km s?1 for all lines.
The wind region in this object, defined by the excess line broadening or
residual velocity dispersions, lies along the galaxy minor axis in the direction of
the well-known radio jet. Bulk velocities of the gas near the galaxy nucleus are
mostly systemic with ?v50 wind? ?10 km s?1 for all lines (see Figure 4.11a.1 -
160
Cen A
(a) ?v50 ?v50
v50 Data v50 Model ?v50 (?W1? > 25) (?W1? > 50)
100 100
49 49 49
?43? 01? 00?? 60 60 27 27 27
?? 20 20 5 5 5?15
-19 -19 -15 -15 -15
?30?? -37 -37 -37
-60 -60
100 100
49 49 49
?43? 01? 00?? 60 60 27 27 27
?? 20 20 5 5 5?15
-19 -19 -15 -15 -15
?30?? -37 -37 -37
-60 -60
100 100
49 49 49
?43? 01? 00?? 60 60 27 27 27
?? 20 20 5 5 5?15
-19 -19 -15 -15 -15
?30?? -37 -37 -37
-60 -60
100 100
49 49 49
?43? 01? 00?? 60 60 27 27 27
20 20 5 5 5
?15??
-19 -19 -15 -15 -15
?30?? -37 -37 184 pc -37
-60 -60
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
(b) W1? Data W1? Model ?W1? ?W1? > 25 ?W1? > 50
283 283 100 100 100
?43? 01? 00?? 216 216 38 38 38
?? 150 150 -22 -22 -22?15
84 84 -84 -84 -84
?30??
17 17 -145 -145 -145
283 283 100 100 100
?43? 01? 00?? 216 216 38 38 38
150 150 -22 -22 -22
?15??
84 84 -84 -84 -84
?30??
17 17 -145 -145 -145
283 283 100 100 100
?43? 01? 00?? 216 216 38 38 38
?? 150 150 -22 -22 -22?15
84 84 -84 -84 -84
?30??
17 17 -145 -145 -145
283 283 100 100 100
?43? 01? 00?? 216 216 38 38 38
?? 150 150 -22 -22 -22?15
84 84 -84 -84 -84
?30?? 184 pc
17 17 -145 -145 -145
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
Figure 4.11 Cen A: Results from modeling the disk velocity field with 3DBarolo.
Symbols and plots are the same as those in Figure 4.8.
161
[[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633 [[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633
[OI] 63 [OIII] 88 [OI] 145 [CII] 158
Figure 4.12 Cen A: PACS contours overlaid on 8.4 GHz image from Hardcastle et al.
(2003).
Cen A
Mwind KEwind Mwind KEwind
(?W1? > 25) (?W1? > 25) (?W1? > 50) (?W1? > 50)
39.73 13.90
?43? 00? 40??
31.88 11.14
?43? 01? 00?? 24.03 8.37
?? 16.17 5.60?20
8.32 2.83
?? Mtot: 1814.68 M KEtot: 434.73 ergs Mtot: ? KE? tot
: ?
40
0.46 0.07
39.73 13.90
?43? 00? 40??
31.88 11.14
?43? 01? 00?? 24.03 8.37
16.17 5.60
?20??
8.32 2.83
?? Mtot: 157.17 M KEtot: 27.17 ergs Mtot: ? KEtot: ??40
0.46 0.07
0.78 0.15
?43? 00? 40??
0.65 0.12
?43? 01? 00?? 0.52 0.10
?? 0.39 0.07?20
0.26 0.05
?? Mtot: 74.73 M KEtot: 11.80 ergs Mtot: ? KEtot: ??40
0.13 0.02
0.78 0.15
?43? 00? 40??
0.65 0.12
?43? 01? 00?? 0.52 0.10
?? 0.39 0.07?20
0.26 0.05
? ??
Mtot: 76.97 M KEtot: 12.28 ergs Mtot: ? KEtot: ?
40
0.13 0.02
0.04 0.04 0.03 0.03
?43? 00? 40??
0.03 0.03 0.02 0.02
?43? 01? 00?? 0.03 0.03 0.02 0.02
?20??
0.02 0.02 0.01 0.01
0.01 0.01 0.01 0.01
?? Mtot: 5.22 M KEtot: 2.17 ergs Mtot: 1.16 M KE : 1.09 ergs? tot40
0.01 0.01 0.01 0.01
30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s 30s 28s 13h 25m 26s
Figure 4.13 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in Figure 4.10.
162
?W1? ?v50
[[OOIIIIII]] 8888 SSoolluuttiioonn DD SSoolluuttiioonn CC SSoolluuttiioonn BB SSoolluuttiioonn AA
Figure 4.11a.4). In [O III] 88 , there is a compact feature ? 15??NE of the nu-
cleus with v50 residuals ? ?30 km s?1 in the E and ? +45 km s?1 in the W. This
feature is not seen in [C II] 158 and shows little bulk motion in either [O I] 63 or
[O I] 145 (?v ?150 wind? +10 km s ).
The residual velocity dispersions in the wind region (Figure 4.11b.1 - Fig-
ure 4.11b.4) near the galaxy nucleus for [O III] 88 and [C II] 158 are spatially com-
pact with ?W1?wind . 30 km s?1 . Also near the nucleus, the residual velocity
dispersions in [O I] 63 and [O I] 145 are elongated towards the NW in the direction
of the jet with ?W ?11?wind? 30?50 km s . The wind region observed in [O I] 145 is
the most extended along the major axis compared to the wind region mapped by the
other three emission lines. The largest residual velocity dispersions in the compact
feature NE of the nucleus is ?W1?wind? 90 km s?1 in [O III] 88 , while moderate
velocity dispersions of ? 30 km s?1 are seen in [O I] 63 and [O I] 145 .
In [O I] 145 an elongated feature is seen S of the galaxy nucleus with ?v50 wind?
25 km s?1 and ?W1?wind? 95 km s?1 . It is unlikely that this feature is a component
of the outflowing gas as it spatially coincides with the edge of the putative warped
disk seen in CO, Pa-?, and the MIR (Espada et al., 2009; Marconi et al., 2000b;
Quillen et al., 2006; Radomski et al., 2008).
Figure 4.12 shows PACS contours of ?v50 wind (top row) and ?W1?wind (bottom
row) overlaid on the 8.4 GHz images of Cen A from Hardcastle et al. (2003). NE of
the nucleus, within r . 1 kpc, a compact feature is seen in [O I] 63 , [O III] 88 , and
[O I] 145 . This feature is spatially coincident with a complex of bright emission
knots seen in the radio and X-ray and which are known to be associated with the
163
outflow (e.g. Hardcastle et al., 2003; Kraft et al., 2002).
For PDR Solution A (see Figure 4.13), the largest neutral outflow masses
(? 4? 105M) are located to the SW of the nucleus in a thin structure elongated
from NW to SE. This is in striking contrast to PDR Solution B where the largest
neutral masses (? 4?103M) are compact and localized around the galaxy nucleus.
Figure 4.13 also shows a compact region of ionized gas around the nucleus that
contains ?50 M and a structure to the NE with mass ?30 M. The largest KE
of 9? 1048 erg lies in a concentrated region to the NE with more moderate energies
of ? 2 ? 1048 centered close to the nucleus. This region of large energy spatially
corresponds to knot ?A? found in Kraft et al. (2002).
4.9.3 Circinus
The v50 maps of Circinus (see Column 1 of Figure 4.14a) range in velocity
between ?100 km s?1 and confirm data at other wavelengths that the N side of the
disk is approaching us and the S side of the disk is receding from us consistent with
observations in H? (Elmouttie et al., 1998b) and in CO (Zschaechner et al., 2016).
Circinus lacks a significant detection of an outflow in any of the fine-structure
lines. Although the [O III] 88 residuals NW of the nucleus lie within the ioniza-
tion cone, the residuals also overlap with a molecular spiral arm seen in CO (Izumi
et al., 2018). Lying on the PACS FOV and spatially coincident with the molec-
ular spiral arm, it is difficult to attribute this residual line width to the outflow.
[O I] 145 residuals in the wind lie SE of the nucleus and are more likely associated
164
Circinus
?v
(a) 50
?v50
v50 Data v50 Model ?v50 (?W1? > 25) (?W1? > 50)
? ? ? ?? 128 12865 20 00 76 76 76
70 70
?15??
41 41 41
12 12 5 5 5
?30??
-45 -45 -29 -29 -29
?45?? -103 -103 -64 -64 -64
? ? ? ?? 128 12865 20 00 76 76 76
70 70
?? 41 41 41?15
12 12 5 5 5
?30??
-45 -45 -29 -29 -29
?45?? -103 -103 -64 -64 -64
? 128 12865? 20? 00?? 76 76 76
70 70
?? 41 41 41?15
12 12 5 5 5
?30??
-45 -45 -29 -29 -29
?45?? -103 -103 -64 -64 -64
?65? 20? 00?? 128 128 76 76 76
70 70
?? 41 41 41?15
12 12 5 5 5
?30??
-45 -45 -29 -29 -29
?45??
204 pc
-103 -103 -64 -64 -64
12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s
(b) W1? Data W1? Model ?W1? ?W1? > 25 ?W1? > 50
? ? ? 255 255 81 81 8165 20 00??
195 195 23 23 23
?15??
135 135 -34 -34 -34
?30??
75 75 -92 -92 -92
?45?? 15 15 -150 -150 -150
? 255 255 81 81 8165? 20? 00??
195 195 23 23 23
?15??
135 135 -34 -34 -34
?30??
75 75 -92 -92 -92
?45?? 15 15 -150 -150 -150
? ? ? ?? 255 255 81 81 8165 20 00
195 195 23 23 23
?15??
135 135 -34 -34 -34
?30??
75 75 -92 -92 -92
?45?? 15 15 -150 -150 -150
? ? 255 255 81 81 8165 20? 00??
195 195 23 23 23
?15??
135 135 -34 -34 -34
?30??
75 75 -92 -92 -92
204 pc
?45?? 15 15 -150 -150 -150
12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s
Figure 4.14 Circinus: Results from modeling the disk velocity field with 3DBarolo.
Symbols, units, and plots are the same as those in Figure 4.8.
165
[[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633 [[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633
Circinus
Mwind KEwind Mwind KEwind
(?W1? > 25) (?W1? > 25) (?W1? > 50) (?W1? > 50)
16.87 4.76
?65? 20? 00?? 13.55 3.82
10.23 2.88
?20??
6.90 1.93
?40??
3.58 0.99
Mtot: ? KEtot: ? Mtot: ? KEtot: ?
0.26 0.05
16.87 4.76
?65? 20? 00?? 13.55 3.82
10.23 2.88
?20??
6.90 1.93
?40??
3.58 0.99
Mtot: ? KEtot: ? Mtot: ? KEtot: ?
0.26 0.05
1.78 1.39
?65? 20? 00?? 1.44 1.11
1.11 0.84
?20??
0.77 0.57
?40??
0.43 0.29
Mtot: ? KEtot: ? Mtot: ? KEtot: ?
0.09 0.02
1.78 1.39
?65? 20? 00?? 1.44 1.11
1.11 0.84
?20??
0.77 0.57
?40??
0.43 0.29
Mtot: ? KEtot: ? Mtot: ? KEtot: ?
0.09 0.02
0.18 0.18 0.08 0.08
?65? 20? 00?? 0.15 0.15 0.06 0.06
0.12 0.12 0.05 0.05
?20??
0.08 0.08 0.04 0.04
?40??
0.05 0.05 0.03 0.03
Mtot: ? KEtot: ? Mtot: ? KEtot: ?
0.01 0.01 0.02 0.02
12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s 12s 14h 13m 08s
Figure 4.15 Circinus does not have a discernible outflow in either the neutral or
ionized gas.
166
[[OOIIIIII]] 8888 SSoolluuttiioonn DD SSoolluuttiioonn CC SSoolluuttiioonn BB SSoolluuttiioonn AA
with the circumnuclear starburst ring (see Figure 5 of Mingo et al., 2012).
4.9.4 NGC 253
For NGC 253, the v50 maps (see Column 1 of Figure 4.16a) range in velocity
from ? ?80 km s?1 to ? 130 km s?1 for all lines. The E side of the disk is approach-
ing us while the W side of the disk is receding from us, consistent with observations
in H? (Hlavacek-Larrondo et al., 2011).
In Figure 4.16(a, b), the wind traced by [O I] 63 is a column oriented along the
galaxy minor axis. The column is approximately ? 5??wide to the NE of the nucleus.
That width increases as distance from the nucleus increases, with maximum widths
of ? 15?? and ? 30?? for the regions N and S of the galaxy center, respectively. The
largest v (? 40 km s?150 ) are located in the S portion of the column while the v50 in
the N portion of the column are nominal (? 0? 5 km s?1 ). This suggests that the
S cone traced by [O I] 63 is receding but there is little (if any) bulk motion of the
gas in the N cone along the line of sight. The velocity dispersion across the column
is nearly uniform with W1?? 20 km s?1 .
The wind in [O III] 88 lies in two distinct regions to the NW and SE of the
nucleus. The NW feature with diameter ? 15?? is located ? 10?? from the galaxy
center while the W edge of the ? 20?? ? 15?? SE feature is adjacent to the nucleus.
Figure 4.16a shows the largest approaching velocities (? ?40 km s?1 ) are con-
centrated in the NW feature. The largest velocity dispersions (? 200 km s?1 )
are prominent in the NW feature while more moderate velocity dispersions (. 90
167
NGC 253
(a) ?v50 ?v50v50 Data v50 Model ?v50 (?W1? > 25) (?W1? > 50)
?25? 17? 00?? 130 130 69 69 69
79 79 40 40 40
?15??
28 28 11 11 11
?30?? -22 -22 -17 -17 -17
-73 -73 -46 -46 -46
?25? 17? 00?? 130 130 69 69 69
79 79 40 40 40
?15??
28 28 11 11 11
?30?? -22 -22 -17 -17 -17
-73 -73 -46 -46 -46
?25? 17? 00?? 130 130 69 69 69
79 79 40 40 40
?15??
28 28 11 11 11
?30?? -22 -22 -17 -17 -17
-73 -73 -46 -46 -46
?25? 17? 00?? 130 130 69 69 69
79 79 40 40 40
?15??
28 28 11 11 11
?30?? -22 -22 -17 -17 -17
189 pc
-73 -73 -46 -46 -46
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
(b) W1? Data W1? Model ?W1? ?W1? > 25 ?W1? > 50
?25? 17? 00?? 414 414 211 211 211
337 337 147 147 147
?15??
260 260 83 83 83
?30?? 183 183 19 19 19
106 106 -43 -43 -43
?25? 17? 00?? 414 414 211 211 211
337 337 147 147 147
?15??
260 260 83 83 83
?30?? 183 183 19 19 19
106 106 -43 -43 -43
?25? 17? 00?? 414 414 211 211 211
337 337 147 147 147
?15??
260 260 83 83 83
?30?? 183 183 19 19 19
106 106 -43 -43 -43
?25? 17? 00?? 414 414 211 211 211
337 337 147 147 147
?15??
260 260 83 83 83
?30?? 183 183 19 19 19
189 pc
106 106 -43 -43 -43
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
Figure 4.16 NGC 253: Results from modeling the disk velocity field with 3DBarolo.
Symbols, units, and plots are the same as those in Figure 4.8.
168
[[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633 [[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633
[OI] 63 [OIII] 88 [OI] 145 [CII] 158
(a)                              (b) (c) (d)
(e) (f) (g) (h)
Figure 4.17 NGC 253: PACS contours (yellow), contours of receding and approach-
ing CO outflow (magenta and blue, respectively) from Bolatto et al. (2013). Com-
posite image from Heesen et al. (2011) which shows H? in red from Westmoquette
et al. (2011), ?20 cm continuum (green), and soft X-ray in blue from Hardcastle et
al. (2010).
km s?1 ) are found in the SE feature.
The wind traced by [O I] 145 is concentrated to the SE of the nucleus with
receding velocities 50 & v50 & 0 km s?1 (see Figure 4.16a) and velocity dispersions
W1?. 115 km s?1 (see Figure 4.16b). The feature is ? 20?? in diameter on the E side
while the W side tapers to the SW for ? 8??. As seen in [O III] 88, the W edge of the
[O I] 145 SW feature lies adjacent to the nucleus. Interestingly, not only does this
SW feature contain similar velocity dispersions (? 115 km s?1 ) in [O III] 88 and
in [O I] 145 , but the extent and location of those velocity dispersions are spatially
coincident in the emission lines. No feature is seen in the NW in [O I] 145 .
Figure 4.16a shows that the majority of [C II] 158 is approaching us with the
largest velocities (v50? ?45) located NW of the nucleus. The velocity dispersions
of [C II] 158 are moderate (? 40? 80 km s?1 ; see Figure 4.16b) in the NW and SE
169
?W1? ?v50
NGC 253
Mwind KEwind Mwind KEwind
(?W1? > 25) (?W1? > 25) (?W1? > 50) (?W1? > 50)
125.25 40.83 55.21 29.25
?25? 17? 00?? 100.30 32.68 44.28 23.45
?? 75.35 24.54 33.34 17.65?15
50.40 16.39 22.41 11.86
?30??
25.45 8.24 11.48 6.06
? ?? Mtot: 10611.05 M KEtot: 3586.39 ergs M45 tot: 2362.56 M KEtot: 1286.40 ergs
0.50 0.10 0.55 0.26
125.25 40.83 55.21 29.25
?25? 17? 00?? 100.30 32.68 44.28 23.45
?? 75.35 24.54 33.34 17.65?15
50.40 16.39 22.41 11.86
?30??
25.45 8.24 11.48 6.06
? ?? Mtot: 1092.11 M45  KEtot: 357.49 ergs Mtot: 161.71 M KEtot: 93.31 ergs
0.50 0.10 0.55 0.26
2.01 0.49 0.75 0.33
?25? 17? 00?? 1.64 0.40 0.63 0.28
?? 1.27 0.31 0.52 0.23?15
0.90 0.22 0.40 0.18
?30??
0.53 0.13 0.28 0.13
? ?? Mtot: 387.03 M KEtot: 121.75 ergs M45 tot: 57.22 M KEtot: 32.38 ergs
0.16 0.04 0.16 0.09
2.01 0.49 0.75 0.33
?25? 17? 00?? 1.64 0.40 0.63 0.28
?? 1.27 0.31 0.52 0.23?15
0.90 0.22 0.40 0.18
?30??
0.53 0.13 0.28 0.13
? ?? Mtot: 400.28 M KEtot: 125.78 ergs Mtot: 59.39 M KEtot: 33.55 ergs45
0.16 0.04 0.16 0.09
0.27 0.27 0.25 0.25
?25? 17? 00?? 0.22 0.22 0.20 0.20
?? 0.17 0.17 0.15 0.15?15
0.11 0.11 0.10 0.10
?30??
0.06 0.06 0.05 0.05
? ?? Mtot: 40.03 M KEtot: 32.31 ergs Mtot: 26.18 M KEtot: 28.31 ergs45
0.01 0.01 0.01 0.01
34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s 34s 0h 47m 32s
Figure 4.18 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in Figure 4.10.
170
[[OOIIIIII]] 8888 SSoolluuttiioonn DD SSoolluuttiioonn CC SSoolluuttiioonn BB SSoolluuttiioonn AA
features. The length of the SE feature is detected along the entire S edge of the FOV
while the width of the feature tapers from the largest width of ? 10?? and decreases
towards the SE. The NW feature is roughly 15??? 10?? and lies along the edge of the
N FOV. Unlike the [O I] 63 and [O I] 145 the SE feature traced by [C II] 158 does
not lie adjacent to the galaxy center.
The greatest amount of ionized gas is found in the SE lobe with masses at
? 350 Mnear the galaxy center and then decreasing in the SE direction with
increasing radius from the nucleus. The ionized gas masses in the NW lobe are
considerably smaller than in the southern lobe, with masses ranging from ? 10 to
? 80 M. We see that the largest energies (? 2 ? 1049 ergs) in the ionized wind
are located in long, thin structure in the SE lobe. We see that the moderate KE are
more extended in the SE lobe.
4.9.5 NGC 1068
The v50 maps of NGC 1068 (see Column 1 in Figure 4.19a) indicate that the
galaxy disk is receding in the W and approaching in the E. However, the bulk mo-
tions of the wind, as determined by the v50 residuals (Column 4 in Figure 4.19a),
are oriented in the opposite direction of the disk rotation with approaching ve-
locities in the W and receding velocities in the E. In [O I] 63 , the largest ap-
proaching ?v50 wind velocities (? ?85 km s?1 ) are in a compact feature just W
of the nucleus while moderate receding velocities are in the E (? 35 km s?1 ).
[O III] 88 ?v50 wind residuals show moderate approaching values in the W (? ?20
171
NGC 1068
(a) ?v50 ?v50
v50 Data v50 Model ?v50 (?W1? > 25) (?W1? > 50)
?0? 00? 30?? 157 157 93 93 93
?40?? 89 89 51 51 51
? ?? 21 21 8 8 850
-46 -46 -33 -33 -33
?0? 01? 00??
-114 -114 -76 -76 -76
?0? 00? 30?? 157 157 93 93 93
?40?? 89 89 51 51 51
? ?? 21 21 8 8 850
-46 -46 -33 -33 -33
?0? 01? 00??
-114 -114 -76 -76 -76
?0? 00? 30?? 157 157 93 93 93
?40?? 89 89 51 51 51
? ?? 21 21 8 8 850
-46 -46 -33 -33 -33
?0? 01? 00??
-114 -114 -76 -76 -76
?0? 00? 30?? 157 157 93 93 93
?40?? 89 89 51 51 51
? 21 21 8 8 850??
-46 -46 -33 -33 -33
?0? 01? 00?? 698 pc
-114 -114 -76 -76 -76
42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s
(b)
W1? Data W1? Model ?W1? ?W1? > 25 ?W1? > 50
?0? 00? 30?? 454 454 205 205 205
?40?? 364 364 133 133 133
? 274 274 60 60 6050??
184 184 -12 -12 -12
?0? 01? 00??
93 93 -85 -85 -85
?0? 00? 30?? 454 454 205 205 205
?40?? 364 364 133 133 133
? ?? 274 274 60 60 6050
? ? ?? 184 184
-12 -12 -12
?0 01 00
93 93 -85 -85 -85
?0? 00? 30?? 454 454 205 205 205
?40?? 364 364 133 133 133
?50?? 274 274 60 60 60
184 184 -12 -12 -12
?0? 01? 00??
93 93 -85 -85 -85
?0? 00? 30?? 454 454 205 205 205
?40?? 364 364 133 133 133
? ?? 274 274 60 60 6050
184 184 -12 -12 -12
?0? 01? 00?? 698 pc
93 93 -85 -85 -85
42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s 42s 41s 2h 42m 40s
Figure 4.19 NGC 1068: Results from modeling the disk velocity field with
3DBarolo. Symbols, units, and plots are the same as those in Figure 4.8.
172
[[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633 [[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[OOII]] 6633
Figure 4.20 NGC 1068: PACS contours overlaid on 349 GHz continuum and CO(3-2)
residual mean-velocity field from Garc??a-Burillo et al. (2014).
173
NGC 1068
Mwind KEwind Mwind KEwind
(?W1? > 25) (?W1? > 25) (?W1? > 50) (?W1? > 50)
858.12 513.98
?0? 00? 30?? 689.84 411.53
?? 521.57 309.09?45
353.30 206.64
?0? 01? 00??
185.03 104.20
Mtot: ? KEtot: ? Mtot: ? KE?? tot
: ?
?15
16.76 1.76
858.12 513.98
?0? 00? 30?? 689.84 411.53
521.57 309.09
?45??
353.30 206.64
?0? 01? 00??
185.03 104.20
Mtot: ? KEtot: ? Mtot: ? KE?? tot
: ?
?15
16.76 1.76
82.20 50.15
?0? 00? 30?? 66.45 40.20
50.69 30.26
?45??
34.93 20.32
?0? 01? 00??
19.17 10.38
M
?? tot
: ? KEtot: ? Mtot: ? KEtot: ?
?15
3.42 0.44
82.20 50.15
?0? 00? 30?? 66.45 40.20
?45??
50.69 30.26
34.93 20.32
?0? 01? 00??
19.17 10.38
Mtot: ? KEtot: ? Mtot: ? KE?? tot
: ?
?15
3.42 0.44
3.27 3.27 3.27 3.27
?0? 00? 30?? 2.66 2.66 2.67 2.67
?? 2.05 2.05 2.06 2.06?45
1.44 1.44 1.45 1.45
?0? 01? 00??
0.83 0.83 0.84 0.84
Mtot: 895.16 M KEtot: 2075.06 ergs Mtot: 548.77 M KEtot: 1900.36 ergs
?15??
0.22 0.22 0.24 0.24
42s 2h 42m 40s 42s 2h 42m 40s 42s 2h 42m 40s 42s 2h 42m 40s
Figure 4.21 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in Figure 4.10.
174
[[OOIIIIII]] 8888 SSoolluuttiioonn DD SSoolluuttiioonn CC SSoolluuttiioonn BB SSoolluuttiioonn AA
km s?1 ) and fast receding velocities in the E (? 65 km s?1 ). [O I] 145 and [C II] 158
show mostly systemic velocities in the W (? ?5 km s?1 ). However, [O I] 145 shows
fast receding velocities in the E (? 90 km s?1 ) while [C II] 158 shows moderate
receding velocities of ? 30 km s?1 .
The largest residual velocity dispersions in the wind (?W1?wind? 200 km s?1 ;
see Figure 4.19b) are detected in [O III] 88 and are in a central compact region that
is elongated towards the NE along the direction of the nuclear bar. This region
is spatially coincident with the molecular outflow seen in CO(3-2) (Garc??a-Burillo
et al., 2014). It is also spatially coincident with H? and [O III] 5007 line emission
which extends NE of the nucleus (Cecil et al., 1990; Veilleux et al., 2003). In the
NLR, the deprojected intrinsic ionized gas velocities associated with optical [NII]
emission are ? 1500 km s?1 with respect to systemic (Cecil et al., 1990).
In [O I] 63, [O I] 145, and [C II] 158, there are two knots which flank the galaxy
nucleus. The W knot contains the larger ?W1?wind residuals with ?W1?wind 170
km s?1 in [O I] 63 and ?W ?11?wind? 100 km s in [O I] 145. The E knot contains
?W1?wind residuals of ? 120 km s?1 and ? 70 km s?1 for [O I] 63 and [O I] 145, re-
spectively. In [C II] 158, the knots have similar dispersion residuals with ?W1?wind?
70 km s?1 . The ?W1?wind knots seen in [O I] 63, [O I] 145, and [C II] 158 are most
likely non-circular motions due to the presence of NGC 1068?s nuclear bar and spiral
arms (features which are not accounted for in our simple disk velocity field model),
as opposed to an outflow.
The ionized gas mass follows the morphological structure of the [O III] 88 flux
(i.e. an ellipse spanning from the NE to the SW) with the greatest masses (? 4000
175
M) concentrated around the galaxy center. The largest energies are concentrated
around the nucleus, while there may be some structure following the spiral arms.
4.9.6 NGC 3079
Column 1 of Figure 4.22a shows the [C II] 158 v50 map of NGC 3079. The
range of velocities are between ? ?120 and ? +160 km s?1 and indicate that the
gas on the N side is approaching us while the gas on the S side is receding from us,
consistent with observations of H? (Veilleux et al., 1999) and H2O maser spots in
the disk (Yamauchi et al., 2004).
The [C II] 158 v50 residuals in the wind region east of the nucleus (Figure 4.22a,
Column 4) show redshifted velocities in the N and blueshifted velocities in the S.
W of the nucleus, most of the redshifted velocities lie close to the galaxy plane
while blueshifted velocities begin to appear further below the plane at ? 10??.
The [C II] 158 ?W1?wind map (Figure 4.22b, Column 4) shows an elongated fea-
ture ? 10?? in length just S of the nucleus with ?W1?wind? 100 km s?1 . A more
extended structure lies E to W around the elongated feature and has ?W1?wind? 75
km s?1 .
The contours of these [C II] 158 residuals are overlaid on the 1.4 GHz image
from Sebastian et al. (2019) (see the top row of Figure 4.23), where the double-lobed
radio morphology of the wind is outlined by bright filaments. At the base of the W
lobe, where the filaments are brightest, ?v50 wind appears to trace the lobe?s edges.
The velocities along the southern edge are receding with ?v ?150 wind? 35 km s and
176
NGC 3079
(a) ?v50 ?v50
v50 Data v50 Model ?v50 (?W1? > 25) (?W1? > 50)
153 153 74 74 74
55? 41? 00??
80 80 776 pc36 36 36
45?? 8 8 -1 -1 -1
-63 -63 -39 -39 -39
55? 40? 30??
-136 -136 -77 -77 -77
10h 02m 00s 10h 02m 00s 10h 02m 00s 10h 02m 00s 10h 02m 00s
(b) W1? Data W1? Model ?W1? ?W1? > 25 ?W1? > 50
370 370 107 107 107
55? 41? 00??
309 309 55 55 776 pc 55
45?? 247 247 3 3 3
186 186 -47 -47 -47
55? 40? 30??
124 124 -99 -99 -99
10h 02m 00s10h 01m 57s 10h 02m 00s10h 01m 57s 10h 02m 00s10h 01m 57s 10h 02m 00s10h 01m 57s 10h 02m 00s10h 01m 57s
Figure 4.22 NGC 3079: Results from modeling the disk velocity field with
3DBarolo. Symbols, units, and plots are the same as those in Figure 4.8.
the velocities along the northern edge are approaching with ?v50 wind? ?15 km s?1 .
The middle row of Figure 4.23 shows the same contours as above, but overlaid
on the continuum subtracted H?+ [N II] image from Cecil et al. (2001) where only
the eastern bubble is visible (the western lobe lies behind the disk of galaxy and is
therefore extinguished). In the E lobe, there is a lack of radio emission along the
southern edge of the bubble, extending out to ? 1 kpc above the disk. However,
there is optical emission along this edge where approaching velocities in [C II] 158 are
?v50 wind? ?35 km s?1 and where the largest velocities of ?W1?wind (? 100 km s?1 ).
Cecil et al. (2001) has shown that this filament aligns with the axis of the jet ob-
served at 8 GHz (Trotter et al., 1998). The larger dispersion residuals observed here
are most likely due to the presence of this jet interacting with the ISM of the galaxy
disk.
At ? 1 kpc, the optical filament appears to disperse at the location where the
177
[[CCIIII]] 115588 [[CCIIII]] 115588
Figure 4.23 Top row: PACS contours of the [C II] 158 v50 and W1? residuals in the
wind in NGC 3079 overlaid on 1.4GHz observations from Sebastian et al. (2019).
Middle row: PACS contours of the [C II] 158 v50 and W1? residuals in the wind in
NGC 3079 overlaid on H?+ [N II] image from Cecil et al. (2001). Bottom row:
Chandra image (blue) + HST (red and green).
178
radio emission begins and extends ? 1 kpc to the top of the bubble. The ?v50 wind
along this filament is ? ?20 km s?1 . The base of the northern edge is seen in both
optical and radio where receding velocities are ?v ?150 wind? 45 km s .
The bottom row of Figure 4.23 shows the [C II] 158 ?W1?wind contours over-
laid on the HST image from above (red and green) and the Chandra data fromCecil
et al. (2001). Notice that in the E, the (soft) X-ray filaments spatially correlate with
the optical filaments, while in the W, the (soft) X-ray emission fills the second lobe,
but lacks any of the filamentous structures seen at 1.4 GHz.
4.9.7 NGC 4945
Column 1 of Figure 4.24a shows the v50 maps of NGC 4945 which range
in velocity from ? ?90 km s?1 to ? +100 km s?1 for [O III] 88 , [O I] 145 , and
[C II] 158 . As discussed in Section 4.8.3, the [O I] 63 emission in NGC 4945 shows
significant self-absorption along the galaxy major axis and is excluded from this
analysis. The v50 maps indicate that the NE part of the disk is receding from us and
the SW part of the disk is approaching us, consistent with previous observations
(Venturi et al., 2017).
The maps of the residuals in the wind in Column 4 of Figure 4.24 show in
[O III] 88 a SE to NW collimated structure aligned with NGC 4945?s biconical out-
flow. The presence of the SE lobe is clearly seen in the [O I] 145 wind residuals, with
perhaps some detection of the inner NW lobe. No outflow is detected in [C II] 158 .
In Figure 4.25, the contours of the wind residuals (magenta lines) are overlaid
179
NGC 4945
?v
(a) 50
?v50
v50 Data v50 Model ?v50 (?W1? > 25) (?W1? > 50)
119 119 86 86 86
?49? 27? 45??
68 68 43 43 43
?49? 28? 00??
17 17 0 0 0
?15?? -33 -33 -42 -42 -42
-84 -84 -85 -85 -85
119 119 86 86 86
?49? 27? 45??
68 68 43 43 43
?49? 28? 00??
17 17 0 0 0
?15?? -33 -33 -42 -42 -42
-84 -84 -85 -85 -85
119 119 86 86 86
?49? 27? 45??
68 68 43 43 43
?49? 28? 00??
17 17 0 0 0
?15?? -33 -33 -42 -42 -42
184 pc
-84 -84 -85 -85 -85
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
(b) W1? Data W1? Model ?W1? ?W1? > 25 ?W1? > 50
?49? 27? 45?? 505 505 278 278 278
? ? ?? 404 404 178 178 178?49 28 00
303 303 78 78 78
?15??
202 202 -21 -21 -21
101 101 -120 -120 -120
?49? 27? 45?? 505 505 278 278 278
? ? ?? 404 404 178 178 178?49 28 00
303 303 78 78 78
?15??
202 202 -21 -21 -21
101 101 -120 -120 -120
?49? 27? 45?? 505 505 278 278 278
? ? ?? 404 404 178 178 178?49 28 00
303 303 78 78 78
?15??
202 202 -21 -21 -21
184 pc
101 101 -120 -120 -120
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.24 NGC 4945: Results from modeling the disk velocity field with
3DBarolo. Symbols, units, and plots are the same as those in Figure 4.8.
180
[[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888 [[CCIIII]] 115588 [[OOII]] 114455 [[OOIIIIII]] 8888
Figure 4.25 Top row: PACS contours overlaid on the [NII] residual velocity field of
NGC 4945 from Venturi et al. (2017). Bottom row: PACS contours overlaid on the
[NII] W70 of NGC 4945 from Venturi et al. (2017).
NGC 4945
Mwind KEwind Mwind KEwind
(?W1? > 25) (?W1? > 25) (?W1? > 50) (?W1? > 50)
56.09 7.71
?49? 27? 45?? 44.93 6.18
?49? 28? 00?? 33.77 4.65
? 22.62 3.1215??
11.46 1.58
?30?? Mtot: ? KEtot: ? Mtot: ? KEtot: ?
0.30 0.05
0.66 0.11
?49? 27? 45?? 0.56 0.09
?49? 28? 00?? 0.47 0.07
? ?? 0.38 0.0615
0.29 0.04
?30?? Mtot: ? KEtot: ? Mtot: ? KEtot: ?
0.20 0.03
0.19 0.17 0.09 0.17
?49? 27? 45?? 0.15 0.14 0.07 0.14
?49? 28? 00?? 0.11 0.10 0.06 0.10
? ?? 0.08 0.07 0.04 0.0715
0.04 0.04 0.02 0.04
?30?? Mtot: 36.37 M KEtot: 22.16 ergs Mtot: 7.29 M KEtot: 13.00 ergs
0.01 0.00 0.01 0.00
30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s 30s 28s 13h 05m 26s
Figure 4.26 Mass and KE in the wind derived from the PDR solutions and the
[O III] 88 flux. Symbols, units, and plots are the same as those in Figure 4.10.
181
[[OOIIIIII]] 8888 SSoolluuttiioonn CC SSoolluuttiioonn AA
on the optical [N II] 6583 A? images from Venturi et al. (2017) (top row) and the
X-ray images from Marinucci et al. (2012) (bottom row). Residual velocities and
dispersions in the ionized gas as traced by [O III] 88 are largest along the edges of
the outflow lobes, this is most evident in Figure 4.25(a,c), with receding velocities
up to ?v ? +95 km s?150 wind and up to ?v50 wind? +50 km s?1 for the SE and NW
lobes, respectively. [O III] 88 residual dispersions at the base of the SE lobe are in
the range ?W1?wind? 100 km s?1 to ? 160 km s?1 , while ?W1?wind ranges from
? 90? 100 km s?1 on the edges of the NW lobe.
The [O I] 145 v50 wind residuals (see Column 1 of Figure 4.24a) are located
mostly below the plane of the galaxy. In the optical, the SE lobe only appears at
? 15?? from the nucleus, but in [O I] 145 , the strongest feature of the SE lobe begins
to emerge at ? 10??, demonstrating the advantage of the FIR to penetrate further
through the galaxy disk. The [O I] 145 feature in the SE lobe is approaching with
?v50 wind? ?90 km s?1 and contains large dispersions in the range of ?W1?wind?
250? 280 km s?1 .
The NW outflow cone in NGC 4945 opens towards us and lies in front of the
galactic disk, while the SE cone is oriented in the opposite geometry (Venturi et al.,
2017) and is partially obstructed by foreground dust located in the galactic plane
(Heckman et al., 1990; Koornneef, 1993). The kinematics of the SE lobe as seen
in [O III] 88 and [O I] 145 are consistent with the MUSE [N II] observations from
Venturi et al. (2017). The center of the SE lobe is approaching us (seen in [O I] 145 )
and the edges are receding (seen in [O III] 88 ). The ?v50 wind of the ionized gas in the
NW lobe however, only indicates that the outflow is receding from us. Venturi et al.
182
(2017) suggests that the edges of the NW lobe are approaching and the inner region
of the NW lobe is receding, however we do not detect any approaching velocities in
the NW lobe.
There is no significant detection of the neutral gas in the outflow of NGC 4945
(see Figure 4.26), but there is a prominent detection of the ionized gas. The largest
ionized masses are concentrated around the galaxy nucleus while the largest KE in
the outflow lie S of the nucleus in a ? 20?? long structure elongated E to W. The
most eastern part of this structure spatially coincides with the SE outflow cone.
Some western portions of this structure may also be attributed to the outflow, but
the most W edge of the structure may be contamination from disk emission.
4.10 Summary
We have analyzed archival Herschel -PACS data of five fine structure lines
([O III] 88 , [O I] 63 , [N II] 122 , [O I] 145 , and [C II] 158 ) in seven nearby galaxies
(Cen A, Circinus, M 82, NGC 253, NGC 1068, NGC 3079, and NGC 4945). While
the galactic-scale outflows in these objects have been studied extensively at wave-
lengths spanning the entire electromagnetic spectrum, they are relatively unexplored
in the FIR wavelength regime, 55 ? 210 ?m , studied here. Thanks to the unique
combination of angular resolution, sensitivity, and two-dimensional coverage in the
FIR of Herschel PACS, we have been able to spatially and kinematically resolve the
neutral atomic and ionized gas in the outflow of each galaxy in our sample. The
main results are summarized as follows:
183
1. We have derived velocity-integrated fluxes, radial velocities, and 1?? line
width maps of all five atomic fine-structure lines. Radial velocity maps indicate that
the kinematics within the galaxy disks are dominated by rotation. For some objects
(e.g. M 82 and Cen A), the observed line dispersion maps alone can delineate the
collimated outflow from the galaxy disk.
2. Analysis of the emission line ratio maps confirm earlier results in the lit-
erature (e.g. the prominent ionization cone in Circinus, the jet in Cen A, and the
ionized outflow in NGC 1068). These ratios are insightful tools to diagnose the
physical conditions of the ionized and neutral gas within the outflows and near the
AGN or SB regions. Compact shapes of the line ratios near the galaxy center reveal
effects of ionizing radiation (e.g. [O I] 63 / [C II] 158 ).
3. Modeling of the PDRs near the AGN and/or SB and within the outflow
has constrained the physical parameters of the neutral atomic gas, namely the UV
interstellar radiation field, G0, and the hydrogen gas density, nH. In general, the
highest densities are compact structures that surround the galaxy center. Exceptions
are seen in M 82 and NGC 253, where the structures are more extended and irregular
in shape. For the most part, the largest ISRFs are also compact structures centered
on the galaxy nucleus; exceptions are M 82 where the shape of the highest radiation
is extended and elongated in the direction of the galaxy disk, and in NGC 4945
where the extended ISRF is elongated to the E. Additionally, although the shape of
G0 in NGC 253 is compact and spherical, it is off center to the W. Two objects show
obvious structures in G0 and/or nHwhich are spatially coincident with the known
outflow. Higher densities and radiation are seen elongated NE in the direction of
184
the jet in Cen A. For NGC 4945, the highest densities are found not around the
galaxy center, but are located SE of the nucleus in the S outflow.
4. We have modeled the gas velocity field of the galaxy disks via a ?tilted-ring?
model. Subsequently, the excess residual between the observed and modeled 1-? line
widths was used to define the spatial location of the wind (e.g. regions where the
1-? line width residuals were > 25 km s?1 ).
5. We have derived properties of the outflow such as bulk motions, masses,
and kinetic energies based on our definition of a wind. We find that excess line
widths are a better indicator of outflow compared to excess radial velocities.
6. We have compared our results with other wavelength data and find that
our definition of a wind results in features (morphological and kinematic) consistent
with the literature. For example, bulk motions in the wind region of M 82 indicate
that the N cone is receding from us and the S cone is approaching us, in agreement
with the literature. Inspection of the excess line widths inside the wind region show
that larger line width residuals coincide with known features in an outflow. For
example, the largest excess line widths in NGC 3079 spatially coincide with the
known radio jet and an optical filament E of the nucleus. Morphology of the wind
is also consistent with the literature. M 82 shows line width residuals extending SE
to NW in a collimated feature which spatially coincides with a known SiO chimney.
7. We have demonstrated the advantage of the FIR wavelength range over
that of the optical and UV when exploring SB or AGN regions. This is most
obvious in NGC 3079 where the outflow W of the nucleus lies behind the disk and
is extinguished in the optical, but the superbubble is detected in [C II] 158.
185
8. For completeness, we also present the results of the analysis of the molecular
gas traced by OH 119 in Appendix B.
We have pushed the boundaries of the capabilities of Herschel -PACS in this
work, but higher resolution observations of AGN and SB regions are necessary to
clearly disentangle the outflow from the galaxy disk. JWST will be able to observe
out to ?28 ?m but with better than ?1?? angular resolution, and therefore out
to distances an order of magnitude larger (?100 Mpc; z ? 0.03). There are many
ionized, neutral, and molecular gas diagnostics in the MIR and this will allow studies
of the multi-phase winds in exquisite details. Moreover, line ratios in the MIR, such
as [O IV] 25.6 / [Ne II] 12.8 and [Ne V] 14.3 / [Ne II] 12.8 will nicely complement
the FIR, providing powerful diagnostics to discriminate AGNs from star formation
activity (e.g. Veilleux et al., 2009).
186
Chapter 5: Summary and Future Work
The goal of this thesis was to study and explore the cool component of galactic-
scale winds in nearby SBs and AGNs. This research is based on archival FIR and
MIR spectroscopic data obtained with Herschel?PACS and Spitzer -IRS, respec-
tively. In this chapter, we briefly summarize the main results of the work discussed
in this thesis.
5.1 Main Results
In Chapter 2 we presented the results of our analysis of spectroscopic data
of OH at 119 ?m obtained with Herschel?PACS of 52 local (d < 50 Mpc) BAT
AGN selected from the very hard X-ray (14 ? 195 keV) 58 month Swift?BAT
Survey. These data were combined with Herschel?PACS observations of 43 nearby
(z < 0.3) ULIRG/QSOs from (Veilleux et al., 2013). The inclusion of the BAT
AGN sample with the ULIRG/QSO sample extended the range of AGN properties
(luminosities, star formation rates, and stellar masses) to lower values by 1 ? 2
orders of magnitude compared to the ULIRG/QSO sample alone. We measure the
frequency of occurrence of the wind in the BAT AGN. We also look for correlations
between the properties of the wind and that of the host galaxy.
187
Evidence for molecular outflows as traced by OH at 119 ?m, was seen in only
four of the 52 BAT AGN. This corresponds to a 24% outflow detection rate in the
BAT AGN where the search for outflows was possible (i.e. 12 objects were seen
in pure absorption and 5 were seen with absorption+emission composite profiles).
Evidence for molecular inflows is seen in seven objects, corresponding to a ? 40%
inflow detection rate in the BAT AGN.
When the BAT AGN are combined with the ULIRG/QSOs, the total sample
covers a range of ? 3 dex in SFR and AGN luminosity and ? 2 dex in stellar
mass. The combined sample shows positive trends between outflow velocities and
the host galaxy properties: stellar mass, SFR, and AGN luminosity. The correlation
between wind velocity and AGN luminosity is the strongest of these three. Our
results strongly suggest that at higher AGN luminosities (log(LAGN/L& 11.5) the
AGN dominates over star formation in driving the outflow. At lower luminosities
the AGN may not have the energetics required to drive fast molecular outflows, but
stellar processes could dominate if the AGN were weak or absent. It is clear that
SF plays a crucial role in driving winds, but the presence of an AGN is needed to
drive the fastest winds in our sample.
In Chapter 3 we presented the analysis of Spitzer?IRS observations of OH
at 35 ?m in 15 nearby (z . 0.06) U/LIRGs. The estimation of molecular out-
flow properties, such as mass, from radiative?transfer models, (such as those in
Gonza?lez-Alfonso et al., 2017), are dependent on the estimation of the OH?to?H
abundance ratio, XOH. Thus, constraining the value of this parameter will result in
more accurate predictions when employed. Therefore, we exploited the fact that the
188
ground?state OH absorption feature at 35 ?m is optically thinner than the other
OH doublets in the FIR to provide an independent constraint on the estimation of
XOH. We computed the OH column densities in our sample and compared them to
the hydrogen column density for a typical optical depth at 35 ?m of ? 0.5. Assum-
ing a gas?to?dust ratio of ? 125, we find a mean of XOH = 1.01? 0.15? 10?6 for
our sample. We attempt to verify that the radiative transfer models from Gonza?lez-
Alfonso et al. (2017) predict a realistic OH 35 line profile by comparing the shape of
the predicted line profile with the shape of the line profile fitted to observed data.
The agreement between the two is generally very good in terms of overall strength
(equivalent width) of the feature, and this adds some confidence in the energetics
derived from these models.
In Chapter 4 we presented the analysis of archival Herschel -PACS data of five
atomic fine structure lines ([O I] 63, [O III] 88, [N II] 122, [O I] 145, and [C II] 158)
in seven nearby galaxies (M 82, Cen A, Circinus, NGC 253, NGC 1068, NGC 3079,
and NGC 4945). Included is also the preliminary analysis of OH 119 in these objects.
We derived velocity?integrated fluxes, radial velocities, and 1-? line width
maps of the fine structure lines in each object. For some objects (e.g. M 82) the
observed line dispersion maps alone can delineate the collimated outflow from the
galaxy disk. We computed emission line ratios and detect well known features in
some of the objects (e.g. the ionization cone in Circinus). We employ PDR models
to estimate the strength of the radiation field and the hydrogen density in our
objects. For some objects higher densities and radiation field strengths are spatially
coincident with the known outflow (e.g. the jet in Cen A). We model the gas
189
velocity fields in our sample with 3DBarolo and use the excess residual between
the observed and modeled line widths to define the spatial location of the wind.
Once the wind is defined, residual radial velocities can trace bulk motions of the
outflow. For example, in M 82, for all atomic lines, the N cone is receding from us
and the S cone is approaching us. The significant advantage of the FIR over optical
or UV wavelengths was underscored while defining the wind location. In particular
for NGC 3079, the FIR captures the W superbubble which lies behind the disk and
is extinguished in the optical. We also estimate masses and kinetic energies for the
neutral and ionized gas in the wind for objects with a wind detection. We also
compare the results of our analysis with other wavelength data and find remarkable
consistency in the comparisons. For example, the largest residual line widths in NGC
3079 spatially coincide with the known radio jet indicating that the larger residuals
are due to the presence of the jet interacting with the ISM of the galaxy disk. The
preliminary analysis of OH 119 in our sample shows unambiguous extended outflows
in NGC 253 in the form of P?Cygni profiles and unambiguous extended inflows in
Circinus and NGC 4945 in the form of inverted P?Cygni profiles.
5.2 Future Work
We have pushed the boundaries of the capabilities of Herschel -PACS in Chap-
ter 4, but higher resolution observations of AGN and SB regions are necessary to
clearly disentangle the outflow from the galaxy disk. The high resolution (? 1??) and
sensitivity of the James Webb Space Telescope (JWST) will be able to resolve the
190
SB and/or AGN activity inside the central spaxel of our PACS observations, thus
providing the opportunity to study the energy mechanisms of winds in great detail.
In the MIR alone, there are many ionized, neutral, and molecular gas diagnostics
that will allow for studies of multiphase winds and the MIR range of JWST will also
be a nice compliment to the FIR since the combination of these wavelength ranges
will expand the possible line ratios such as [O IV] 25.6 / [O III] 88 which has been
proposed as a powerful diagnostic to discriminate AGNs from SF activity.
Also of particular interest will be SPace Infrared telescope for Cosmology and
Astrophysics (SPICA; Swinyard et al., 2009) whose unprecedented spectroscopic
sensitivity in the FIR will be able shed light on the role of molecular outflows at
high z (currently, only a few detections have been reported). At higher redshifts, the
star formation rate density and AGN activity increase which points to an increasing
importance of feedback processes earlier in the Universe. SPICA will be able to
measure outflow properties of distant AGN with unprecedented accuracy and depth,
and provide us with a greater understanding of galaxy formation and evolution.
191
Appendix A: Spitzer MIR SPECTRA
10.0 CenA Circinus 10.00 ESO 005-G004
100.0
1.00
1.0
10.0
0.10
0.1 1.0 0.01
10.0 10.0 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
10.0 IC 5063 MCG-05-23-016 1.0 MCG-06-30-015
1.0
1.0
0.1 0.1 0.1
10.0 10.0 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
1.00 Mrk 18 1.00 NGC 1052 NGC 1365
10.0
0.10 0.10
1.0
0.01 0.01 0.1
10.00 10.00 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
NGC 2110 1.00 NGC 2655 10.00 NGC 2992
1.00
1.0
0.10
0.10
0.1 0.01 0.01
10.0 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
Figure A.1 Mid-infrared (5-37?m) spectra used to measure S9.7?m . The dashed
line is the continuum calculated from the cubic spline interpolation fitted to the
pivot points shown as black dots. Red dots show fcont(9.7?m) (located on dashed
continuum line) and fobs(9.7?m) (located on the solid black line or the observed flux
density). The blue line shows the integration range used to calculate the flux and
total equivalent width of the 9.7 ?m silicate feature (see Table 2.4 and Table 2.5).
192
Jy Jy Jy Jy
Jy Jy Jy
Jy
Jy Jy Jy Jy
10.00 NGC 3079 NGC 3081 NGC 3227
1.00
1.00
1.0
0.10
0.10
0.01 0.01 0.1
10.00 10.00 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
10.0 NGC 3281 10.0 NGC 3783 NGC 4051
1.0
1.0 1.0
0.1 0.1 0.1
10.0 10.0 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
NGC 4102 10.0 NGC 4151 1.0 NGC 4258
10.0
1.0
1.0
0.1
0.1 0.1
10.0 10.0 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
10.0 NGC 4388 NGC 4395 1.0 NGC 4593
0.100
1.0
0.010
0.1
0.001 0.1
10.0 10.000 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
Figure A.2 (Continued)
193
Jy Jy Jy Jy
Jy Jy Jy Jy
Jy Jy Jy Jy
1.000 NGC 4939 1.00 NGC 4941 100.00 NGC 4945
10.00
0.100
0.10 1.00
0.010
0.10
0.001 0.01 0.01
10.000 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
NGC 5273 10.0 NGC 5506
NGC 5728
1.00
0.10
1.0
0.10
0.01
10.00 10.0 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
1.00 NGC 5899 NGC 6221 10.0 NGC 6300
10.00
1.00
0.10 1.0
0.10
0.01 0.1
10.00 10.00 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
1.00 NGC 6814 10.00 NGC 7172 1.00 NGC 7213
1.00
0.10
0.10
0.10
0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
Figure A.3 (Continued)
194
Jy Jy Jy Jy
Jy Jy Jy Jy
Jy Jy Jy Jy
NGC 7314 10.00 NGC 747910.00 NGC 7582
10.0
1.00 1.00
1.0
0.10 0.10
0.01 0.01 0.1
10.00 10.00 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
F00509+1225 F01572+0009 1.000 F05024-1941
1.00
1.0
0.100
0.10
0.010
0.001
0.1 0.01
10.0 10.00 10.000
Wavelength [?m] Wavelength [?m] Wavelength [?m]
10.0 F05189-2524 10.000 07251-0248 1.0 F07598+6508
1.000
1.0 0.100
0.010
0.1 0.001 0.1
10.0 10.000 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
10.00 F08572+3915 10.00 09022-3615 10.00 F09320+6134
1.00 1.00 1.00
0.10 0.10 0.10
0.01 0.01 0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
Figure A.4 (Continued)
195
Jy Jy Jy Jy
Jy JyJy Jy
Jy
Jy Jy
Jy
10.00 F10565+2448 1.0 F11119+3257 F12072-0444
1.00
1.00
0.10
0.10 0.1
0.01 0.01
10.00 10.0 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
10.000 F12112+0305 100.00 F12243-0036 1.0 F12265+0219
1.000 10.00
0.100 1.00
0.010 0.10
0.001 0.01 0.1
10.000 10.00 10.0
Wavelength [?m] Wavelength [?m] Wavelength [?m]
F12540+5708 13120-5453 F13428+5608
10.00
10.00
10.00
1.00
1.00
0.10 0.10
1.00
0.01 0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
F13451+1232 10.00 F14348-1447 10.00 F14378-3651
1.00
1.00 1.00
0.10 0.10 0.10
0.01 0.01
0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
Figure A.5 (Continued)
196
Jy Jy Jy Jy
Jy Jy Jy Jy
Jy JyJy
Jy
F14394+5332 F15206+3342 10.00 F15250+3609
1.00
1.00
1.00
0.10
0.10 0.10
0.01 0.01
0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
F15327+2340 F15462-0450 F16504+0228
10.00
10.00 1.00
1.00
1.00
0.10
0.10
0.10
0.01 0.01 0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
F17208-0014 F19297-0406 19542+1110
10.00 10.00
1.00
1.00 1.00
0.10
0.10
0.10
0.01
0.01
0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
10.00 F20551-4250 F22491-1808 F23128-5919
1.00
1.00
1.00
0.10
0.10
0.10
0.01
0.01 0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
Figure A.6 (Continued)
197
Jy Jy Jy Jy
Jy Jy Jy Jy
Jy Jy Jy Jy
F23365+3604 1.00 PG 1351+640 1.00 PG 1440+356
1.00
0.10 0.10 0.10
0.01
0.01 0.01
10.00 10.00 10.00
Wavelength [?m] Wavelength [?m] Wavelength [?m]
1.00 PG 1613+658 1.00 PG 2130+099
0.10
0.10
0.01
10.00 10.00
Wavelength [?m] Wavelength [?m]
Figure A.7 (Continued)
198
Jy
Jy
Jy
Jy
Jy
Appendix B: Results on OH 119 ?m
PACS observations of the OH 119.233, 119.441 ?m doublet were retrieved
and reduced in the same manner as the atomic FIR data (see Section 4.4). The
PACS resolution at 119 ?m is ? 270 km s?1 . Details of the OH observations are
in Table B.1. To reduce noise, the OH data have been smoothed with a Gaussian
kernel of width 0.05 ?m (this step was excluded for the spectra in NGC 253). For
each spaxel, a first-order spline was fit to the continuum and subtracted from the
spectra. Profile fitting of the OH doublet followed a similar procedure from Veilleux
et al. (2013) and Stone et al. (2016). The OH profile is modeled using four Gaussian
components (two for each line of the doublet). The separation between the two
lines of the doublet was set to 0.208 ?m in the rest frame (? 520 km s?1 ) and the
amplitude and standard deviation were fixed to be the same for each component in
the doublet.
We characterize the OH line profile properties in the same manner as the
atomic lines (e.g. v50 and W1? ) and present here our preliminary results of the
spectral analysis of OH 119 for each object. Four figures are presented for each object
(see Figure B.1). The top left shows the PACS IFU footprint on the sky (black lines
and squares) and the outline of the ionized wind traced by [O III] 88 (white line).
199
Table B.1. PACS OH 119 ?m Observations
Object Name ObsId texp (s) Proposal ID
M 82 1342232257 958 OT1 shaileyd 1
Cen A 1342225989 976 OT1 shaileyd 1
Circinus 1342225147 958 OT1 shaileyd 1
NGC 253 1342237604 976 OT1 shaileyd 1
NGC 1068 1342203128 3330 KPGT esturm 1
NGC 3079 1342221391 8045 DDT esturm 4
NGC 4945 1342247792 958 OT1 shaileyd 1
Note. ? Column 1: Galaxy Name. Column 2: Obser-
vation Id. Column 3: Exposure time. Column 4: Proposal
ID.
Top right shows the spline fits to the continuum while bottom left shows the line
profile fits to the continuum subtracted spectrum. Finally, bottom right shows maps
of the line profile properties for the absorption and/or emission components fitted
to the spectra. The 1-? line width maps have not been corrected for instrumental
broadening.
When pure emission and/or pure absorption (e.g. M 82 and NGC 1068) is
observed in the OH 119 line profile, the rotation of the galaxy disk is discernible in
the v50 maps (see Figure B.1 and Figure B.5). The rotation is consistent with the
results from the analysis of the ionized and neutral gas traced by the fine structure
lines. However, when there is a combination of absorption and emission components
in an OH line profile, the physical interpretation of v50 is not so clear.
For all of objects, there is no detection of an outflow in the central spaxel.
Only Circinus (see Figure B.3) shows a wind in the central spaxel in the form of an
unambiguous inflow (e.g. an inverted P-Cygni profile is observed). Cen A shows no
200
discernible detection of a wind in OH 119. NGC 1068 is seen in pure emission with
no detection of an outflow.
The largest line widths (? 500 km s?1 ) in M 82 (see Figure B.1) appear in
pure absorption in spaxels (3,0) and (2,3). The absorption lines have blue-shifted
wings and are spatially coincident with the ionized outflow detected in [O III] 88,
suggesting there is also outflow in the molecular gas.
NGC 253 shows detections of unambiguous outflow outside of the central
spaxel (see Figure B.4) which extend N, S, and NW of the nucleus. The P-Cygni
profiles lie in good spatial agreement with the ionized outflow traced by [O III] 88 .
NGC 4945 (see Figure B.7) shows pure absorption above and below the disk at
the edges of the PACS FOV. Unambiguous inflow is seen mostly S of the nucleus in
spaxels (2,1), (4,2), and (4,3). These spaxels are spatially coincident with the edges
of the ionized outflow traced by [O III] 88 . There is also unambiguous inflow W of
the nucleus in spaxel (3,3), but the spaxel lies outside of the ionized outflow. There
is also a possible detection of a blue absorption wing E of the nucleus in spaxel (1,1).
201
M82 1342232257 OH 119 ?m
37 48
17 20
4x4 11536 4x3 4x2 4x1 4x0
46
69?41'15" 1635 19110
4x1 4x0 4434 15 18
105 42
33
17
100 40
14
32
4x2 3x1 3x0 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
33
3x4 160 3x3 3x2 26 3x1 3x018
125
00" 4x3
32
25
150
17
3x2 2x1 31 1204x4 2x0 24
140 16
30
115 23
3x3 29 130 15
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
2x2 2003x4 1x1
50
60
1x0 95.023 2x4 2x3 195 2x2 2x1 2x0
58 92.5 48
190
40'45" 2x3 22 90.056 185 46
87.5
2x4 1x2 21 1800x1 540x0 4485.0175
52 82.5
20 42
1x3 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 300076 240
1x4 0x2
14 1x4 30 1x3 1x2 1x1 115 1x0
74 230
13 29 110
30" 0x3 72 22028
12 105
70
0x4 27 210 10011
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
11.0 31
22 61 35
10.5 0x4 0x3 0x2 0x1 0x0
30 60 34
10.0 21
9.5 59 33
15" 20
29
9.0 58 32
8.5
19 28 57 31
8.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
9h55m57s 54s 51s 48s Velocity [km s?1]
M82 1342232257 OH 119 ?m
Total Flux [10?17 Wm?2] Abs Total Flux [10?17 Wm?2] Emi Total Flux [10?17 Wm?2]
M82 1342232257 OH 119 ?m 3.51 -1.87 3.51
4 4 4
0.0 0 0
4x4 4x3 4x2 4x1 4x0 -16.23 -20.54 3.170.0 0.0
?0.5 ?1 3 3 3
?5
? ?0.2? 21.0 ?0.2 -35.97 -39.20 2.83
?3 ?0.4
? 2 2 21.5 ?10
?4 ?0.6 ?0.4 -55.71 -57.86 2.48
?2.0
? ?5 1 1 115
? ?0.82.5 ?0.6 -75.45 -76.52 2.14
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0 0 0 0 00.0
0.0 0.03x4 3x3 ?2 3x2 3x1 3x0 -95.19 -95.19 1.80
?5 ?0.5
? ?0.20.2 ?4 4 3 2 1 0 4 3 2 1 0 4 3 2 1 0
? ?10 ?1.0
?0.4
0.4 ?6
? ?8?0.6 15 ?
?0.6
1.5 v50,abs [km/s] v50,emi [km/s]
?10 ?0.8
?0.8 ?20 ?2.0 119 61
?12
?1.0 4 4
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.4 85 150 0
0.0 0
2x4 2x3 2x2 2x1 2x0 3 3
0.3
? ?20.5 ?5 ?1 52 -31
0.2
?1.0 ?4 2 2
0.1 ?10 ?2
? 19 -77?1.5 6
1 1
0.0
? ?15 ? ?32.0 8
-13 -123
?0.1
? 0 02000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.0
0.0 0 0.0 -47 -169
0.8
0.8 1x4 1x3 1x2 1x1 1x0?0.5 ?5 ?2.5 4 3 2 1 0 4 3 2 1 0
0.6
0.6 ?1.0 ?10 ?5.0
0.4
0.4 ?1.5 ?7.5 W1?,abs [km/s] W1?,emi [km/s]?15
0.2
0.2 ?2.0 ? 515 846? 10.020
0.0 4 4
0.0 ?2.5 ?12.5
?25
470 744
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.8 3 3
0.8 0.0 0.0
0x4 0.8 0x3 0x2 0x1 0x0 426 642
0.6
0.6 ?0.5 ?0.5
0.6 2 2
0.4
0.4 ?1.0 382 541
0.4 ?1.0
0.2 ?1.5 1 1
0.2 0.2
?1.5 338 439
0.0 ?2.0
0.0 0.0 0 0
?2.0 ?2.5
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 294 338
Velocity [km s?1] 4 3 2 1 0 4 3 2 1 0
Figure B.1 Top left: PACS IFU footprint of OH 119 (black squares) overlaid
on the 22 ?m WISE image of M 82. The white contour marks the spatial ex-
tent of the [O III] 88 wind. The black contour outlines the PACS footprint of the
[O III] 88 observation. The black cross marks the adopted galaxy center and the
blue line marks the galaxy major axis. Top right: Spline fits to the OH 119 contin-
uum (blue dashed lines). Black lines are the observed data. Magenta areas indicate
the regions used to fit the continuum. Blue dots mark the pivot points used to
fit the spline. Bottom left: Line profile fitting results of the continuum-subtracted
spectra. Solid blue lines indicate gaussian absorption components. Solid red lines
indicate gaussian emission components. Vertical dashed blue (red) lines mark the
v16 , v50 , and v84 velocities in absorption (emission). Bottom right: Top row, from
left to right shows the total velocity-integrated flux of the fitted OH line profiles,
the total flux in the absorption components only, and the total flux in the emission
components only. Middle row shows v50 for the absorption (left) and emission com-
ponents (right). Bottom row shows the 1?? line widths of the absorption (left) and
emission (right) components.
202
Flux [Jy]
Flux [Jy]
CENA 1342225989 OH 119 ?m
5.5 5.5 4.5
5.0
0x0 0x1 5.5 0x2 5.0 0x3 0x4
5.0 4.0
4.5
0x4 5.0 4.5 3.5
4.5
4.0
4.0
4.5 3.0
-43?00'45" 4.03.50x3 3.54.0 2.5
3.0 3.5
1x4 3.03.5 2.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
8.0 10.0 12.5
0x2 8.51x3 1x0 9.5 1x1 6.512.0 1x2 1x3 1x47.5
8.0
2x4 9.0 11.5 6.07.0
7.5
8.5 11.0 5.5
6.5
10.5 7.0
8.0
01'00" 0x1 1x2 2x3 5.06.0 10.0 6.57.5
0x0 3x4 4.59.5 6.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
28
15.0 7.5
9.0 11.0
2x2 3x3 2x0 14.5 2x1 2x2 2x3 2x427 7.01x1
1x0 4x4
8.5
14.0 10.5
6.5
8.0 13.5 26 10.0
13.0 6.0
7.5
25 9.5
12.5
15" 3x2 4x3 5.57.02x1 9.012.0 24
2x0 5.0?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
6.5
8.0 13.5 9.5 8.0
4x2 3x0 3x1 3x2 3x3 3x46.0 13.0 7.5
7.5 9.0
3x1 12.5
5.5 7.0
3x0 7.0 8.512.0
6.5
5.0
6.5 11.5 8.0
11.0 6.0
6.0 4.5 7.5
30" 4x1 10.5 5.54x0 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
4.0 4.0 6.5
4x0 4x1 5.0 4x2 6.5 4x3 4x4
3.5 3.5 6.0
4.5 6.0
3.0
3.0 5.5
4.0 5.5
2.5 5.0
2.5
3.5 5.0
2.0 4.5
2.0
3.0 4.5
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
13h25m30s 28s 26s Velocity [km s?1]
CENA 1342225989 OH 119 ?m
Total Flux [10?17 Wm?2] Abs Total Flux [10?17 Wm?2] Emi Total Flux [10?17 Wm?2]
CENA 1342225989 OH 119 ?m 0.13 -0.14 0.13
0.15 0 0 0
0.075 0.05
0.10
0x0 0x1 0.05 -0.77 -0.99 0.130.050 0.10 0x2 0x3 0x4
0.05 0.00 1 1 1
0.025 0.05
0.00 -1.68 -1.84 0.12
0.00
0.000 0.00 ?0.05 2 2 2
?0.05
?0.025 ?0.05?0.05 -2.59 -2.70 0.12
?0.10
? ?0.100.050
? 3 3 30.10
? ?0.100.15 ?0.075 -3.49 -3.55 0.11
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.10 0.10
0.05 4 4 4
0.0
1x0 1x1 0.05 1x2 1x3 0.1 1x4
0.00 -4.40 -4.40 0.110.05
? ?0.10.05 0.00 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0.00
0.0
?0.10 ?0.05 ?0.2
?0.05
?0.15 ?0.10 v50,abs [km/s] v50,emi [km/s]
? ?0.10.3
?0.20 ?0.10 ? 284 -870.15
?0.25 ?0.4 ?0.2 0 0
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
208 -126
0.00 0.05
0.0 0.0
2x0 2x1 2x2 2x3 2x4 1 10.0
?0.25 0.00
132 -166
?0.1 ?0.2 ?0.50 ?0.1 ?0.05 2 2
? ?0.75 ?0.100.2 56 -206
?0.4 ?0.2
?1.00 ?0.15 3 3
?0.3 ?1.25 ?0.3 ?0.20
? -19 -2460.6
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ? 4 41000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.1
0.1
0.10 -95 -286
3x0 3x1 0.0 3x2 0.0 3x3 3x4
0.0 0 1 2 3 4 0 1 2 3 4
0.05
?0.2 0.0
?0.1 0.00 ?0.1
? W [km/s] W [km/s]0.05 ?0.4 1?,abs 1?,emi
? ?0.10.2 ?0.2
? 873 2500.10
?0.6
?0.3 0 0?0.15 ?0.3 ?0.2
758 239
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.05 1 1
0.075 0.04 0.10 0.10
4x0 4x1 4x2 4x3 4x4 643 229
0.050 0.02 0.000.05
0.05 2 2
0.025
0.00 0.00 ?0.05 529 218
0.00
0.000 ?0.02 ?0.05 ?0.10 3 3
?0.025
? ?0.05 ?0.100.04 414 208
?0.15
?0.050 ?0.15 4 4
?0.06 ?0.10
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 300 198
Velocity [km s?1] 0 1 2 3 4 0 1 2 3 4
Figure B.2 Cen A. Lines and symbols are the same as those in Figure B.1.
203
Flux [Jy]
Flux [Jy]
CIRCINUS 1342225147 OH 119 ?m
17.5 9.0 4.0
4.0 15.0
0x0 0x1 17.0 0x2 8.5 0x3 0x43.5
3.5 14.5
16.5
8.0
3.0
3.0 14.0 16.0
7.5
15.5 2.5
13.5
0x4 2.5 7.015.0 2.0
13.0
2.0 14.5 6.5
-65?20'00" 0x3 12.5 1.5?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
38 4.5
5.5
0x2 1x0 1x1 1x2 1x3
1x4 4.0
1x4
5.0 33 37 13
1x3 3.54.5 36
0x0 32 120x1 3.04.0
35
1x2 31 2.5112x4 3.5 34 2.0
15" 2x3 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
1x0 967.5 4.5
1x1 2x0 2x1 2x2
17.5 2x3 2x4
2x2 37 94 17.0 4.03x4 7.0 16.5
3.5
6.5 92
3x3 36 16.0 3.0
2x0 6.0
90
15.5
2x1 35 2.53x2 88 15.04x4 5.5 2.0?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
30" 5.04x3 50.0 16.516.56.5 3x0 3x1 3x2 3x3 4.5 3x4
49.5 16.0
3x0 16.0 4.03x1 6.04x2 49.0 15.515.5 3.5
48.5
5.5 15.0
3.0
15.0 48.0
5.0 14.5 2.5
4x0 47.514.54x1 14.0 2.04.5 47.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
6.0 22.5
45" 12.5 4.022.0
5.5 4x0 4x1 4x2 10.0 4x3 4x4
12.0 3.5
21.5
5.0 9.511.5
3.0
21.0
11.0
4.5 9.0 2.5
20.5
10.5
4.0 8.520.0 2.0
10.0
3.5 19.5 8.0 1.5
9.5
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
14h13m15s 12s 09s 06s Velocity [km s?1]
CIRCINUS 1342225147 OH 119 ?m
Total Flux [10?17 Wm?2] Abs Total Flux [10?17 Wm?2] Emi Total Flux [10?17 Wm?2]
CIRCINUS 1342225147 OH 119 ?m 2.77 -0.08 2.77
0 0 0
0.3
0.05 0.0750.25
0x0 0x1 0x2 0.4 0x3 0x4 1.88 -0.75 2.24
0.20 0.00 0.050 1 1 1
0.3 0.2
0.15 ?0.05 0.025 0.99 -1.42 1.71
0.2
0.10 ?0.10 0.000 2 2 2
0.1
0.1
0.05 ?0.15 ?0.025 0.10 -2.09 1.18
0.00 ?0.20 ?0.050 0.0
0.0 3 3 3
?0.05 ?0.075 ?0.1 -0.80 -2.76 0.64
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.0 0.4 4 4 4
0.3
0.6
1x0 1x1 1x2 1x3 0.40.8 1x4
0.3 -1.69 -3.43 0.11
0.2 0.6 0.3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0.2 0.4
0.4 0.2
0.1 0.1 0.2
0.2 0.1 v50,abs [km/s] v50,emi [km/s]
0.0
0.0 0.0 892 3770.0 0.0
?0.1 0 0
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.0 638 2470.5
2x0 0.6 1 10.15 0.4 2x1 2x2 2x3 2x4
0.5 0.4
384 117
0.4
0.10 0.2 0.30.0 2 2
0.2 0.2
0.05
0.0 ? 130 -110.5
0.1 3 3
0.0
0.00
?0.2 ?1.0
0.0 -123 -141
?0.2
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ? 4 42000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.2
0.6 -377 -271
0.20
0.10 3x0 3x1 0.1 3x2 3x3 3x4
0.2 0 1 2 3 4 0 1 2 3 4
0.15
0.0 0.4
0.05 0.10 ?0.1
0.1
W [km/s] W [km/s]
0.05 ?0.2 0.2 1?,abs 1?,emi
0.00
?0.3 0.0 402 574
0.00
?0.4 0.0 0 0
?0.05 ?0.05 ?0.1 361 492
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1 1
0.4 0.5
4x0 4x1 0.3
0.3
4x2 4x3 4x4 321 411
0.2 0.4
0.3 2 2
0.2 0.2
0.3
0.2
0.1 281 330
0.2
0.1 0.1
0.1 3 3
0.1
0.0 241 249
0.0 0.0 0.0
0.0
4 4
? ?0.10.1
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 201 168
Velocity [km s?1] 0 1 2 3 4 0 1 2 3 4
Figure B.3 Circinus. Lines and symbols are the same as those in Figure B.1.
204
Flux [Jy]
Flux [Jy]
NGC253 1342237604 OH 119 ?m
10.0 20.0 17.5
-25?16'45" 9.50x0 19.5 0x1 0x2 17.09.5 0x3 0x4
21
16.5 9.019.0
9.0
18.5 16.0 8.5
20
8.5
18.0 15.5 8.0
8.0 17.5 19 15.0 7.5
0x0 17.0
14.5
7.5 7.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
29 40
130 13
17'00" 0x2 0x3 1x0 1x1 75 1x2 1x3 1x428 380x1 1250x4 127027 36120
1x0
26 115 65 34 11
1x2 1x3 25
110 32
60 10
105
1x1 1x4 24 30?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
550 90 19
2x0 32.0 922x0 2x1 2x2 2x3 18 2x415" 31.5 90 855002x2 2x3 31.0 88 1780
2x1 30.52x4 86 450 16
30.0
84 75
3x0 1529.5
82 400
14
29.0 70
3x2 803x3 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
170
3x1 3x4 10 16 95
4x0 3x0 3x1 3x2 3x3
24 3x4
30" 16015 909
22
4x2 14 150 854x3 84x1 204x4 13 140 80
7
12 75 18
130
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
11 27 28 16
12
4x0 4x1 26 4x2 4x3 4x4
10 15
45" 25 2611 9 1424
24
10 23 13
8
22
22 12
9 7
21
11
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
0h47m35s 34s 33s 32s 31s Velocity [km s?1]
NGC253 1342237604 OH 119 ?m
Total Flux [10?17 Wm?2] Abs Total Flux [10?17 Wm?2] Emi Total Flux [10?17 Wm?2]
NGC253 1342237604 OH 119 ?m 7.53 -0.35 12.38
0.2 0 0 0
0.2 0.2 0.6
0.4 -86.24 -92.54 10.04
0x0 0x1 0x2 0x3 0.1 0x4
0.1
1 1 1
0.0
0.1 0.4
0.0 0.2 -180.01 -184.74 7.69
?0.1
?0.1 2 2 2
0.0 0.2
0.0 ?0.2
?0.2 -273.78 -276.93 5.35
?0.3
? ?0.2 3 3 3? 0.3
0.0
0.1
?0.4 -367.54 -369.12 3.01
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.25
0 4 4 40.0 1
0
1x0 1x1 1x2 1x3 1.00 1x4 -461.31 -461.31 0.66
? ?
0
2
0.5 ?5 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
? 0.75? 14
?1.0 ?10 ? ?2 0.506
v50,abs [km/s] v50,emi [km/s]?8 ?3 0.25
?1.5 ?15 9 321
?10 ?4 0.00
0 0
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
-28 201
0.4 00 2.5 1.5
2x0 2x1 2x2 2x3 2x4 1 1
?25 0.0 1.0
?2 -66 81
0.2
?50 ?2.5 0.5 2 2
?4 ?75 ?5.0 0.0 -103 -390.0
?6 ?100 ?7.5 ? 3 30.5
?0.2 -141 -159?125 ?10.0 ?1.0
?8
?2000 ? 4 41000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
5 1 -179 -279
0.0
1.5
1.5 3x0 3x1 3x2 3x3 3x40 ? 0 1 2 3 4 0 1 2 3 42.5 0
1.0 ?5 ?5.0
1.0
?1
?10 ?7.5 W1?,abs [km/s] W1?,emi [km/s]
0.5
0.5
? ?10.015 ?2 371 605
?12.5 0 0
0.0 0.0
?20 ?3
?15.0 328 523
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.5 1 1
0.3 0.0
4x0 0.6 4x1 4x2 4x3 4x4 285 442
1.0 0
0.2 ?0.5 2 2
0.4
0.5
? 243 3601
0.1 0.2 0.0 ?1.0 3 3
?0.5 ?2 200 279
0.0 0.0 ?1.5
?1.0 4 4
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 158 198
Velocity [km s?1] 0 1 2 3 4 0 1 2 3 4
Figure B.4 NGC 253. Lines and symbols are the same as those in Figure B.1.
205
Flux [Jy]
Flux [Jy]
NGC1068 1342203128 OH 119 ?m
-0?00'15" 7.0 6.56.0 1611.0
4x4 4x3 4x2 6.5 4x1 4x0
5.5 6.0
10.5
15
5.0 6.0
10.0 5.5
4.5 5.5
9.5 5.0
14
4.0
9.0 5.0
4.5
4x4 3.5 8.5 4.513 4.0
4x3 4x1 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 30004x2 11.0 20.5 12.030" 10.5 3x4 3x3 24 3x2 17.5 3x1 3x020.0 11.5
17.0
4x0 10.0 2319.5 11.03x4 3x1 16.59.5 19.0 22 10.53x3 16.03x2 9.0 18.5 10.021
15.5
8.5 18.0 9.5
15.0
3x0 8.0
20
17.5 9.0
2x4 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 30002x1 9.0 14.5 54 20.5
45" 2x3 14.02x2 8.5 2x4 2x3 2x2 21 2x1 20.0 2x052
13.5
8.0
20 19.5
13.0 50
2x0 7.5 19.01x4 1912.5 481x1 7.0 18.512.0 18
1x3 1x2 6.5
46 18.0
11.5
6.0 44 17 17.5
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
11.0
31
19
0x4 1x0 5.0 1x4 10.5 1x3 1x2 1x1 16.5 1x001'00" 300x3 0x1 4.5 10.0 18 16.00x2 29
15.5
4.0 9.5
17 28
9.0 15.0
3.5
0x0 8.5 27 14.5
3.0 16
8.0 14.026
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
4.0 21.5
0x4 5.5 0x3 9.5 0x2 0x1 6.0 0x0
21.0
3.5
15" 5.0 9.0 5.520.5
3.0
8.5
4.5 20.0 5.0
2.5 8.0 19.5
4.0 4.5
2.0 7.5
19.0
3.5 4.0
18.5
7.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
2h42m43s 42s 41s 40s 39s Velocity [km s?1]
NGC1068 1342203128 OH 119 ?m
Total Flux [10?17 Wm?2] Abs Total Flux [10?17 Wm?2] Emi Total Flux [10?17 Wm?2]
NGC1068 1342203128 OH 119 ?m 23.92 23.92
0.6 0.8
4 4 4
0.4
0.4
4x4 0.6 4x3 4x2 4x1 4x0 19.18 19.18
0.6 0.3
0.3
0.4 3 3 3
0.4 0.4
0.2 0.2 14.44 14.44
0.2 0.2 0.1 2 2 2
0.2 0.1
0.0 9.71 9.71
0.0
0.0 0.0 0.0? 1 1 10.1
?0.2 4.97 4.97
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.0 0 0 0
0.8 0.8
3x4 3x3 2.00.8 3x2 3x1 0.6 3x0 0.23 0.23
0.6 0.6
0.6 1.5 4 3 2 1 0 4 3 2 1 0 4 3 2 1 0
0.4
0.4 0.4
0.4 1.0
0.2
0.2 0.2 0.2 v50,abs [km/s] v50,emi [km/s]0.5
199
0.0 0.0 0.00.0 0.0
4 4
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.0 0.8 122
0.6 2.06
2x4 0.8 2x3 2x2 2x1 2x0 3 3
0.6
1.5 45
0.4 0.6 4
0.4 2 2
1.0
0.4
0.2 -30
2 0.2
0.2 0.5 1 1
0.0 0.00.0 -107
0 0.0
? ? 0 02000 1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
1.25
-184
0.3 0.6 0.6
1x4 1x3 1.00 1x2 2.0 1x1 1x0 4 3 2 1 0 4 3 2 1 0
0.2 0.4 0.75 1.5 0.4
0.50 1.0 W1?,abs [km/s] W1?,emi [km/s]0.2
0.1 0.2
0.25
0.5 631
0.0
0.0 0.00 0.0 4 4
0.0
568
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.3 0.5 3 3
0.8 0.15
0x4 0x3 0.4 0x2 0x1 0x0 505
0.05 0.2
0.6 0.10
0.3 2 2
0.1 0.4 0.05 4420.00 0.2
1 1
0.00
0.1 0.2
? 0.00.05 379
0.0 ?0.05
0.0 0 0
?0.1
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 316
Velocity [km s?1] 4 3 2 1 0 4 3 2 1 0
Figure B.5 NGC 1068. Lines and symbols are the same as those in Figure B.1.
206
Flux [Jy]
Flux [Jy]
NGC3079 1342221391 OH 119 ?m
3.0 4.0
1.5
55?41'15" 1.0 0x0 0x1 4.52.5 0x2 0x3 0x43.5
1.0
4.0
0.5 2.0
3.0
0x4 0.5 3.50.0 1.5
2.5
0x3 0.0?0.5 3.01.0
2.0
?0.5
?1.0
0x2 0.5
2.5
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
1x4 2.53.0 6.5 8.5
1.0 1x0 1x1 1x2 1x3 1x4
8.0 2.0
00" 0x0 1x3 2.5 6.00x1 0.5 7.52.0 5.5 1.5
7.0
1x2 0.0 1.5 5.0 1.0
6.5
2x4 ?0.5 1.0 4.5 6.0 0.5
4.0
1x0 2x3
0.5
? 5.51.0 0.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
1x1 4.0 2.01.5 30
10
2x2 2x0 3.5 2x1 2x2 2x3 1.5 2x41.0
28
3x4 0.5 3.0 9 1.0
40'45" 26
2x0 3x3 0.0 2.5 0.524 8
2x1 ?0.5 2.0 0.022
3x2 7?1.0 1.5 20 ?0.5
4x4 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 30004.51.5 1.5
4.0 17
4x3 3x0 3x1 3x2 4.0 3x3 3x43x0 1.0 1.0163.5
3x1 3.50.5
4x2 15 0.53.0 3.0
30" 0.0 14 0.0
2.5 2.5
?0.5 13 ?0.5
4x0 2.0 2.04x1 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 30002.0
4.5 2.5 1.5
4.5
4x0 4x1 4x2 4x3 4x4
1.5 2.04.0 1.0
4.0
1.5
1.0 3.5 0.5
3.5
1.0
0.5 3.0 0.0
3.0
15" 0.52.50.0 ?0.52.5
0.0
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
10h02m00s 01m58s 56s 54s Velocity [km s?1]
NGC3079 1342221391 OH 119 ?m
Total Flux [10?17 Wm?2] Abs Total Flux [10?17 Wm?2] Emi Total Flux [10?17 Wm?2]
NGC3079 1342221391 OH 119 ?m 0.23 -0.12 0.23
0.10 0.025 0 0 00.10
0.02
0x0 0x1 0x2 0.000 0x3 0x4 -5.28 -5.55 0.19
0.05 0.05
0.05 0.00? 1 1 10.025
?0.02 -10.78 -10.99 0.160.00
0.00 ?0.050
0.00
2 2 2
?0.075 ?0.04
?0.05 -16.29 -16.43 0.13
?0.05 ?0.05 ?0.100 ?0.06
3 3 3
?0.10 ?0.125 ?0.08
-21.80 -21.87 0.09
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.050 4 4 40.075 0.0
1x0 0.0 1x1 1x2 0.0 1x3 1x4
0.025 -27.30 -27.30 0.06
0.050
?0.1 ?0.2 ?0.2 0.000 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0.025
?0.025
?0.2
0.000 ?0.4 ?0.4
?0.050 v50,abs [km/s] v50,emi [km/s]
?0.025 ?0.3
?0.6 ?0.6 ?0.075 340 362
?0.050
?0.4 ?0.100
?0.8 0 0
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.2 241 246
0.15 0
2x0 2x1 2x2 0.0 2x3 0.05 2x4 1 1
0.10 0.0
?2 142 131
?0.2 0.00
0.05 2 2
?0.1
? ?0.44 43 15
0.00 ?0.05
?0.6
? 3 3? 0.20.05 ?6 ?0.10
?0.8 -56 -99
?0.10
?2000 ? 4 41000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.100 0.0 0.10 -155 -215
3x0 3x1 3x2 0.0 3x3 3x4
0.05
0.075 0 1 2 3 4 0 1 2 3 4
?0.5
0.05
0.050 ?0.1
0.00
?1.0
0.025 W [km/s] W [km/s]0.00 1?,abs 1?,emi
?0.2
0.000 756 745?0.05 ?1.5
?0.05
? 0 00.025 ?0.3
?2.0 642 629
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
0.10 1 1
0.050
0.05
4x0 0.05 4x1 0.05 4x2 4x3 4x4 528 513
0.025
0.05
0.00 2 2
0.00 0.00 0.000
414 397
0.00 ?0.05 ?0.025
?0.05 ?0.05 3 3
?0.050
?0.05 ?0.10 300 281
?0.10 ?0.10 ?0.075
4 4
?0.15
?0.10 ?0.100
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 186 166
Velocity [km s?1] 0 1 2 3 4 0 1 2 3 4
Figure B.6 NGC 3079. Top left: The white contour marks the spatial extent of the
[C II] 158 wind. Lines and symbols are the same as those in Figure B.1.
207
Flux [Jy]
Flux [Jy]
NGC4945 1342247792 OH 119 ?m
26 21
40
0x0 0x1 0x2 0x3 8 0x4
6
25 20
39
19
5 24 7
0x4 23 38 18
4 6
0x3 22 17-49?27'45" 373 21 16 5
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
0x2 421x4 12 101x0 1x1 160 1x2 41 1x3 1x476
11
1x3 9
10 150
40
0x0 74 80x1 9 39140
1x2 7722x4 8 38130 6
7 70 37
28'00" 2x3 120 5
1x0 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 300052 13
1x1 12 52
2x2 2x0 2x1
550 2x2 2x3 12 2x4
3x4 50 11
50 500
10 48
3x3 10450
48
46 9
2x0 8 4002x1 83x2 46 444x4 350 7
6
15" ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 30004x3 16
17 180
7
3x0 3x0 3x1 3x2 52 3x3 3x43x1 16 144x2 6 160 5015
5
12
14 48
140
4
13
46
4x0 104x1 3 12 120 44
30" ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 30009 19
30 12
4x0 4x1 4x2 22 4x3 4x4
18
11
28
8
17 21
10
26
16
20
7 9
15 24
19 8
14
22
6
13 18
7
?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000 ?3000?2000?1000 0 1000 2000 3000
13h05m30s 28s 26s 24s Velocity [km s?1]
NGC4945 1342247792 OH 119 ?m
Total Flux [10?17 Wm?2] Abs Total Flux [10?17 Wm?2] Emi Total Flux [10?17 Wm?2]
NGC4945 1342247792 OH 119 ?m 17.90 -4.13 17.90
3.0 0 0 0
3.0
1.25 0.5 1.0
0x0 0x1 0x2 2.52.5 0x3 0x4
-129.14 -146.76 14.44
1.00 0.8
0.0 1 1 12.0
2.0
0.75 0.6 -276.18 -289.40 10.98
1.5 ? 1.50.5
0.50 0.4 2 2 2
1.0 1.0
?1.0
0.2 -423.23 -432.04 7.520.25
0.5 0.5
?1.5 3 3 3
0.00 0.0 0.0
0.0
-570.27 -574.67 4.05
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
3.0 0 2.5 4 4 40
2.5 1x0 1x1 1x2 0.0 1x3 1x4 -717.31 -717.31 0.59
2.0
?1
2.0 ?10 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
?0.5 1.5
1.5 ?2
1.0
1.0 ?20
? ?1.0 v3 50,abs [km/s] v50,emi [km/s]
0.5 0.5
224 495
0.0 ?4 ?30 ?1.5 0.0
0 0
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
192 347
0 5
1
2x0 2x1 2x2 2x3 2x4 1 1
3 34
?50 161 199
0 3
2 2 2 2
?100 2
130 51
?1
1 1
? 1150 3 3
?2 0 99 -96
0 ? 0200
? ? 4 42000 1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
5
2.5 0
68 -244
2.5 1
3x0 3x1 3x2 3x3 4 3x4
?10 0 0 1 2 3 4 0 1 2 3 42.0 2.0
1.5 1.5 ?20 ?1
3
?30 ?2 2 W1?,abs [km/s] W1?,emi [km/s]1.0 1.0
0.5 0.5 ?
?3
40 346 4611
? ?4 0 00.0 0.0 50 0
?5 322 410
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000
2.0 1.0 1 1
1.0 2.5
4x0 4x1 1 4x2 4x3 4x4 298 359
0.8 1.5 0.5 2.0
0 2 2
0.6 1.0 ?1 0.0 1.5 274 308
0.4 ?2 1.0
0.5 ? 3 30.5
0.2 ?3 0.5 250 257
0.0 ?1.0
0.0 ?4 0.0 4 4
?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 ?2000 ?1000 0 1000 2000 227 206
Velocity [km s?1] 0 1 2 3 4 0 1 2 3 4
Figure B.7 NGC 4945. Lines and symbols are the same as those in Figure B.1.
208
Flux [Jy]
Flux [Jy]
Bibliography
Abel, N. P., Dudley, C., Fischer, J., Satyapal, S., & van Hoof, P. A. M. 2009, ApJ,
701, 1147
Aguirre, A., Hernquist, L., Schaye, J., et al. 2001, ApJ, 561, 521
Alatalo, K., Blitz, L., Young, L. M., et al. 2011, ApJ, 735, 88
Alexander, D. M., Swinbank, A. M., Smail, I., McDermid, R., & Nesvadba, N. P. H.
2010, in Astronomical Society of the Pacific Conference Series, Vol. 427, Accretion
and Ejection in AGN: a Global View, ed. L. Maraschi, G. Ghisellini, R. Della Ceca,
& F. Tavecchio, 74
Alton, P. B., Davies, J. I., & Bianchi, S. 1999, A&A, 343, 51
Armus, L., Mazzarella, J. M., Evans, A. S., et al. 2009, PASP, 121, 559
Baldry, I. K., Glazebrook, K., Brinkmann, J., et al. 2004, ApJ, 600, 681
Banerji, M., Chapman, S. C., Smail, I., et al. 2011, MNRAS, 418, 1071
Barbosa, F. K. B., Storchi-Bergmann, T., McGregor, P., Vale, T. B., & Rogemar
Riffel, A. 2014, MNRAS, 445, 2353
Barnes, J. E., & Hernquist, L. 1996, ApJ, 471, 115
Bassani, L., Dadina, M., Maiolino, R., et al. 1999, ApJS, 121, 473
Baumgartner, W. H., Mushotzky, R., & Tueller, J. 2011, in Bulletin of the American
Astronomical Society, Vol. 43, American Astronomical Society Meeting Abstracts
#217, #125.01
Bell, E. F., & de Jong, R. S. 2001, ApJ, 550, 212
Benson, A. J., Bower, R. G., Frenk, C. S., et al. 2003, ApJ, 599, 38
Bertone, S., De Lucia, G., & Thomas, P. A. 2007, MNRAS, 379, 1143
Blanco, V. M., Graham, J. A., Lasker, B. M., & Osmer, P. S. 1975, ApJ, 198, L63
209
Bland, J., Taylor, K., & Atherton, P. D. 1987, MNRAS, 228, 595
Blanton, M. R., & Moustakas, J. 2009, ARA&A, 47, 159
Bolatto, A. D., Warren, S. R., Leroy, A. K., et al. 2013, Nature, 499, 450
Bosma, A. 1978, in IAU Symposium, Vol. 77, Structure and Properties of Nearby
Galaxies, ed. E. M. Berkhuijsen & R. Wielebinski, 28
Bower, R. G., Benson, A. J., Malbon, R., et al. 2006, MNRAS, 370, 645
Bradford, C. M., Stacey, G. J., Fischer, J., et al. 1999, in ESA Special Publication,
Vol. 427, The Universe as Seen by ISO, ed. P. Cox & M. Kessler, 861
Bravo-Guerrero, J., & Stevens, I. R. 2017, MNRAS, 467, 3788
Brusa, M., Perna, M., Cresci, G., et al. 2016, A&A, 588, A58
Burns, J. O., Feigelson, E. D., & Schreier, E. J. 1983, ApJ, 273, 128
Calzetti, D., Wu, S.-Y., Hong, S., et al. 2010, ApJ, 714, 1256
Cano-D??az, M., Maiolino, R., Marconi, A., et al. 2012, A&A, 537, L8
Carilli, C. L., & Walter, F. 2013, ARA&A, 51, 105
Carniani, S., Marconi, A., Maiolino, R., et al. 2015, A&A, 580, A102
?. 2016, A&A, 591, A28
Cecil, G., Bland, J., & Tully, R. B. 1990, ApJ, 355, 70
Cecil, G., Bland-Hawthorn, J., & Veilleux, S. 2002, ApJ, 576, 745
Cecil, G., Bland-Hawthorn, J., Veilleux, S., & Filippenko, A. V. 2001, ApJ, 555,
338
Chevalier, R. A., & Clegg, A. W. 1985, Nature, 317, 44
Cicone, C., Brusa, M., Ramos Almeida, C., et al. 2018, Nature Astronomy, 2, 176
Cicone, C., Feruglio, C., Maiolino, R., et al. 2012, A&A, 543, A99
Cicone, C., Maiolino, R., Sturm, E., et al. 2014, A&A, 562, A21
Cicone, C., Maiolino, R., Gallerani, S., et al. 2015, A&A, 574, A14
Cielo, S., Bieri, R., Volonteri, M., Wagner, A. Y., & Dubois, Y. 2018, MNRAS, 477,
1336
Clarke, D. A., Burns, J. O., & Norman, M. L. 1992, ApJ, 395, 444
210
Cohen, M., Megeath, S. T., Hammersley, P. L., Mart??n-Luis, F., & Stauffer, J. 2003,
AJ, 125, 2645
Cole, S., Norberg, P., Baugh, C. M., et al. 2001, MNRAS, 326, 255
Combes, F. 2014, Proceedings of the International Astronomical Union, 10, 182189
Combes, F., & Charmandaris, V. 2000, A&A, 357, 75
Contursi, A., Poglitsch, A., Gra?cia Carpio, J., et al. 2013, A&A, 549, A118
Cooper, J. L., Bicknell, G. V., Sutherland, R. S., & Bland-Hawthorn, J. 2008, ApJ,
674, 157
?. 2009, ApJ, 703, 330
Costa, T., Sijacki, D., & Haehnelt, M. G. 2014, MNRAS, 444, 2355
Crenshaw, D. M., Kraemer, S. B., Boggess, A., et al. 1999, ApJ, 516, 750
Cresci, G., Mainieri, V., Brusa, M., et al. 2015, ApJ, 799, 82
Croton, D. J., Springel, V., White, S. D. M., et al. 2006, MNRAS, 365, 11
Croxall, K. V., Smith, J. D., Wolfire, M. G., et al. 2012, ApJ, 747, 81
Dale, D. A., Helou, G., Brauher, J. R., et al. 2004, ApJ, 604, 565
Dale, D. A., Helou, G., Contursi, A., Silbermann, N. A., & Kolhatkar, S. 2001, ApJ,
549, 215
Darling, J. 2007, in IAU Symposium, Vol. 242, Astrophysical Masers and their
Environments, ed. J. M. Chapman & W. A. Baan, 417?426
Darling, J., & Giovanelli, R. 2006, AJ, 132, 2596
Das, V., Crenshaw, D. M., Kraemer, S. B., & Deo, R. P. 2006, AJ, 132, 620
Dasyra, K. M., Tacconi, L. J., Davies, R. I., et al. 2006a, ApJ, 638, 745
?. 2006b, ApJ, 651, 835
?. 2007, ApJ, 657, 102
de Grijs, R. 2001, Astronomy and Geophysics, 42, 4.12
Destombes, J. L., Marliere, C., Baudry, A., & Brillet, J. 1977, A&A, 60, 55
Di Matteo, T., Springel, V., & Hernquist, L. 2005, Nature, 433, 604
Di Teodoro, E. M., & Fraternali, F. 2015, MNRAS, 451, 3021
Diamond-Stanic, A. M., Moustakas, J., Tremonti, C. A., et al. 2012, ApJ, 755, L26
211
D??az-Santos, T., Armus, L., Charmandaris, V., et al. 2017, ApJ, 846, 32
Dinerstein, H. L., Lester, D. F., & Werner, M. W. 1985, ApJ, 291, 561
Draine, B. T. 2011, Physics of the Interstellar and Intergalactic Medium
Dunn, J. P., Crenshaw, D. M., Kraemer, S. B., & Trippe, M. L. 2008, AJ, 136, 1201
Duric, N., Seaquist, E. R., Crane, P. C., Bignell, R. C., & Davis, L. E. 1983, ApJ,
273, L11
Elbaz, D., Daddi, E., Le Borgne, D., et al. 2007, A&A, 468, 33
Elmouttie, M., Haynes, R. F., Jones, K. L., Sadler, E. M., & Ehle, M. 1998a,
MNRAS, 297, 1202
Elmouttie, M., Koribalski, B., Gordon, S., et al. 1998b, MNRAS, 297, 49
Engelbracht, C. W., Kundurthy, P., Gordon, K. D., et al. 2006, ApJ, 642, L127
Espada, D., Matsushita, S., Peck, A., et al. 2009, ApJ, 695, 116
Fabian, A. C. 2012, ARA&A, 50, 455
Fabian, A. C., Sanders, J. S., Haehnelt, M., Rees, M. J., & Miller, J. M. 2013,
MNRAS, 431, L38
Falstad, N., Gonza?lez-Alfonso, E., Aalto, S., et al. 2015, A&A, 580, A52
Farrah, D., Lebouteiller, V., Spoon, H. W. W., et al. 2013, ApJ, 776, 38
Ferkinhoff, C., Hailey-Dunsheath, S., Nikola, T., et al. 2010, ApJ, 714, L147
Ferkinhoff, C., Brisbin, D., Nikola, T., et al. 2011, ApJ, 740, L29
Ferna?ndez-Ontiveros, J. A., Spinoglio, L., Pereira-Santaella, M., et al. 2016, ApJS,
226, 19
Feruglio, C., Maiolino, R., Piconcelli, E., et al. 2010, A&A, 518, L155
Filippenko, A. V., & Sargent, W. L. W. 1992, AJ, 103, 28
Fischer, J., Abel, N. P., Gonza?lez-Alfonso, E., et al. 2014, ApJ, 795, 117
Fischer, J., Sturm, E., Gonza?lez-Alfonso, E., et al. 2010, A&A, 518, L41
Flower, D. R., Pineau des Forets, G., Field, D., & May, P. W. 1996, MNRAS, 280,
447
Fo?rster Schreiber, N. M., Genzel, R., Newman, S. F., et al. 2014, ApJ, 787, 38
Garc??a-Burillo, S., Mart??n-Pintado, J., Fuente, A., & Neri, R. 2001, ApJ, 563, L27
212
Garc??a-Burillo, S., Combes, F., Usero, A., et al. 2014, A&A, 567, A125
Genzel, R., Fo?rster Schreiber, N. M., Rosario, D., et al. 2014, ApJ, 796, 7
Ginsburg, A., & Mirocha, J. 2011, PySpecKit: Python Spectroscopic Toolkit, As-
trophysics Source Code Library, , , ascl:1109.001
Goicoechea, J. R., & Cernicharo, J. 2002, ApJ, 576, L77
Goicoechea, J. R., Cernicharo, J., Lerate, M. R., et al. 2006, ApJ, 641, L49
Goldsmith, P. F., Y?ld?z, U. A., Langer, W. D., & Pineda, J. L. 2015, ApJ, 814, 133
Gonza?lez-Alfonso, E., & Cernicharo, J. 1999, in ESA Special Publication, Vol. 427,
The Universe as Seen by ISO, ed. P. Cox & M. Kessler, 325
Gonza?lez-Alfonso, E., Fischer, J., Aalto, S., & Falstad, N. 2014a, A&A, 567, A91
Gonza?lez-Alfonso, E., Fischer, J., Gracia?-Carpio, J., et al. 2012, A&A, 541, A4
?. 2014b, A&A, 561, A27
Gonza?lez-Alfonso, E., Fischer, J., Sturm, E., et al. 2015, ApJ, 800, 69
Gonza?lez-Alfonso, E., Fischer, J., Spoon, H. W. W., et al. 2017, ApJ, 836, 11
Goulding, A. D., Alexander, D. M., Bauer, F. E., et al. 2012, ApJ, 755, 5
Greene, J. E., Zakamska, N. L., & Smith, P. S. 2012, ApJ, 746, 86
Griffiths, R. E., Ptak, A., Feigelson, E. D., et al. 2000, Science, 290, 1325
Gu?ltekin, K., Richstone, D. O., Gebhardt, K., et al. 2009, ApJ, 698, 198
Habing, H. J. 1968, Bull. Astron. Inst. Netherlands, 19, 421
Hailey-Dunsheath, S., Nikola, T., Stacey, G. J., et al. 2010, ApJ, 714, L162
Hardcastle, M. J., Worrall, D. M., Kraft, R. P., et al. 2003, ApJ, 593, 169
Harris, G. L. H., Rejkuba, M., & Harris, W. E. 2010, PASA, 27, 457
Harrison, C. M., Alexander, D. M., Mullaney, J. R., & Swinbank, A. M. 2014,
MNRAS, 441, 3306
Harrison, C. M., Alexander, D. M., Swinbank, A. M., et al. 2012, MNRAS, 426,
1073
Harrison, C. M., Alexander, D. M., Mullaney, J. R., et al. 2016, MNRAS, 456, 1195
Hashimoto, T., Laporte, N., Mawatari, K., et al. 2018, Nature, 557, 392
213
Heckman, T. M., Alexandroff, R. M., Borthakur, S., Overzier, R., & Leitherer, C.
2015, ApJ, 809, 147
Heckman, T. M., Armus, L., & Miley, G. K. 1990, ApJS, 74, 833
Heckman, T. M., Lehnert, M. D., Strickland, D. K., & Armus, L. 2000, ApJS, 129,
493
Heckman, T. M., & Thompson, T. A. 2017, arXiv e-prints, arXiv:1701.09062
Heesen, V., Beck, R., Krause, M., & Dettmar, R. J. 2011, A&A, 535, A79
Helou, G., Khan, I. R., Malek, L., & Boehmer, L. 1988, ApJS, 68, 151
Herrera-Camus, R., Bolatto, A., Smith, J. D., et al. 2016, ApJ, 826, 175
Herrera-Camus, R., Sturm, E., Gracia?-Carpio, J., et al. 2018a, ApJ, 861, 94
?. 2018b, ApJ, 861, 95
Higdon, S. J. U., Devost, D., Higdon, J. L., et al. 2004, PASP, 116, 975
Hlavacek-Larrondo, J., Carignan, C., Daigle, O., et al. 2011, MNRAS, 411, 71
Hollenbach, D. J., Takahashi, T., & Tielens, A. G. G. M. 1991, ApJ, 377, 192
Hollenbach, D. J., & Tielens, A. G. G. M. 1997, ARA&A, 35, 179
?. 1999, Reviews of Modern Physics, 71, 173
Hoopes, C. G., Heckman, T. M., Strickland, D. K., et al. 2005, ApJ, 619, L99
Hopkins, P. F., Cox, T. J., Keres?, D., & Hernquist, L. 2008, ApJS, 175, 390
Hopkins, P. F., Somerville, R. S., Hernquist, L., et al. 2006, ApJ, 652, 864
Houck, J. R., Roellig, T. L., van Cleve, J., et al. 2004, ApJS, 154, 18
Hutton, S., Ferreras, I., Wu, K., et al. 2014, MNRAS, 440, 150
Irwin, J., Wiegert, T., Merritt, A., et al. 2019, AJ, 158, 21
Israel, F. P., Gu?sten, R., Meijerink, R., Requena-Torres, M. A., & Stutzki, J. 2017,
A&A, 599, A53
Itoh, T., Done, C., Makishima, K., et al. 2008, PASJ, 60, S251
Izumi, T., Wada, K., Fukushige, R., Hamamura, S., & Kohno, K. 2018, ApJ, 867,
48
Janssen, A. W., Christopher, N., Sturm, E., et al. 2016, ApJ, 822, 43
214
Jarrett, T. H., Chester, T., Cutri, R., Schneider, S. E., & Huchra, J. P. 2003, AJ,
125, 525
Kaufman, M. J., Wolfire, M. G., & Hollenbach, D. J. 2006, ApJ, 644, 283
Kaufman, M. J., Wolfire, M. G., Hollenbach, D. J., & Luhman, M. L. 1999, ApJ,
527, 795
Kennicutt, Jr., R. C. 1998, ApJ, 498, 541
Kimm, T., Somerville, R. S., Yi, S. K., et al. 2009, MNRAS, 394, 1131
King, A. R. 2010, in Astronomical Society of the Pacific Conference Series, Vol. 427,
Accretion and Ejection in AGN: a Global View, ed. L. Maraschi, G. Ghisellini,
R. Della Ceca, & F. Tavecchio, 315
King, A. R., Zubovas, K., & Power, C. 2011, MNRAS, 415, L6
Koornneef, J. 1993, ApJ, 403, 581
Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511
Kormendy, J., & Richstone, D. 1995, ARA&A, 33, 581
Kornei, K. A., Shapley, A. E., Martin, C. L., et al. 2012, ApJ, 758, 135
?. 2013, ApJ, 774, 50
Koss, M., Mushotzky, R., Baumgartner, W., et al. 2013, ApJ, 765, L26
Koss, M., Mushotzky, R., Veilleux, S., et al. 2011, ApJ, 739, 57
Kraft, R. P., Forman, W. R., Jones, C., et al. 2002, ApJ, 569, 54
Krieger, N., Bolatto, A. D., Walter, F., et al. 2019, ApJ, 881, 43
Krips, M., Mart??n, S., Eckart, A., et al. 2011, ApJ, 736, 37
Krug, H. B., Rupke, D. S. N., & Veilleux, S. 2010, ApJ, 708, 1145
Laine, S., & Beck, R. 2008, ApJ, 673, 128
Lehnert, M. D., & Heckman, T. M. 1996, ApJ, 462, 651
Leroy, A. K., Walter, F., Martini, P., et al. 2015, ApJ, 814, 83
Liseau, R., Justtanont, K., & Tielens, A. G. G. M. 2006, A&A, 446, 561
Lutz, D., Sturm, E., Genzel, R., et al. 2003, A&A, 409, 867
Lutz, D., Berta, S., Contursi, A., et al. 2015, ArXiv e-prints, arXiv:1511.02075
Madau, P., Ferrara, A., & Rees, M. J. 2001, ApJ, 555, 92
215
Maiolino, R., Alonso-Herrero, A., Anders, S., et al. 2000, ApJ, 531, 219
Maiolino, R., Gallerani, S., Neri, R., et al. 2012, MNRAS, 425, L66
Malkan, M. A., Gorjian, V., & Tam, R. 1998, ApJS, 117, 25
Marconi, A., & Hunt, L. K. 2003, ApJ, 589, L21
Marconi, A., Moorwood, A. F. M., Origlia, L., & Oliva, E. 1994, The Messenger,
78, 20
Marconi, A., Oliva, E., van der Werf, P. P., et al. 2000a, A&A, 357, 24
Marconi, A., Schreier, E. J., Koekemoer, A., et al. 2000b, ApJ, 528, 276
Marinucci, A., Risaliti, G., Wang, J., et al. 2012, MNRAS, 423, L6
Martin, C. L. 2005, ApJ, 621, 227
?. 2006, ApJ, 647, 222
Martin, C. L., Shapley, A. E., Coil, A. L., et al. 2012, ApJ, 760, 127
?. 2013, ApJ, 770, 41
Martin, C. L., & Soto, K. T. 2016, ApJ, 819, 49
Matsuta, K., Gandhi, P., Dotani, T., et al. 2012, ApJ, 753, 104
McCormick, A., Veilleux, S., Mele?ndez, M., et al. 2018, MNRAS, 477, 699
McElroy, R., Croom, S. M., Pracy, M., et al. 2015, MNRAS, 446, 2186
McKinley, B., Briggs, F., Gaensler, B. M., et al. 2013, MNRAS, 436, 1286
McKinley, B., Tingay, S. J., Carretti, E., et al. 2018, MNRAS, 474, 4056
Meijerink, R., Spaans, M., & Israel, F. P. 2007, A&A, 461, 793
Meijerink, R., Spaans, M., Loenen, A. F., & van der Werf, P. P. 2011, A&A, 525,
A119
Mele?ndez, M., Mushotzky, R. F., Shimizu, T. T., Barger, A. J., & Cowie, L. L.
2014, ApJ, 794, 152
Mele?ndez, M., Veilleux, S., Martin, C., et al. 2015, ApJ, 804, 46
Middelberg, E., Krichbaum, T. P., Roy, A. L., Witzel, A., & Zensus, J. A. 2005,
Astronomical Society of the Pacific Conference Series, Vol. 340, Approaching NGC
3079 with VLBI, ed. J. Romney & M. Reid, 140
Middelberg, E., Roy, A. L., Nagar, N. M., et al. 2004, A&A, 417, 925
216
Mingo, B., Hardcastle, M. J., Croston, J. H., et al. 2012, ApJ, 758, 95
Moorwood, A. F. M., van der Werf, P. P., Kotilainen, J. K., Marconi, A., & Oliva,
E. 1996, A&A, 308, L1
Morganti, R., Frieswijk, W., Oonk, R. J. B., Oosterloo, T., & Tadhunter, C. 2013,
A&A, 552, L4
Morganti, R., Killeen, N. E. B., Ekers, R. D., & Oosterloo, T. A. 1999, MNRAS,
307, 750
Morganti, R., Veilleux, S., Oosterloo, T., Teng, S. H., & Rupke, D. 2016, A&A, 593,
A30
Mushotzky, R. F., Shimizu, T. T., Mele?ndez, M., & Koss, M. 2014, ApJ, 781, L34
Naab, T., & Ostriker, J. P. 2017, ARA&A, 55, 59
Nardini, E., Reeves, J. N., Gofford, J., et al. 2015, Science, 347, 860
Nenkova, M., Ivezic?, Z?., & Elitzur, M. 2002, ApJ, 570, L9
Nenkova, M., Sirocky, M. M., Nikutta, R., Ivezic?, Z?., & Elitzur, M. 2008, ApJ, 685,
160
Nesvadba, N. P. H., Polletta, M., Lehnert, M. D., et al. 2011, MNRAS, 415, 2359
Netzer, H., Lutz, D., Schweitzer, M., et al. 2007, ApJ, 666, 806
Offer, A. R., & van Dishoeck, E. F. 1992, MNRAS, 257, 377
Ott, S. 2010, in Astronomical Society of the Pacific Conference Series, Vol. 434,
Astronomical Data Analysis Software and Systems XIX, ed. Y. Mizumoto, K.-I.
Morita, & M. Ohishi, 139
Parkin, T. J., Wilson, C. D., Schirm, M. R. P., et al. 2014, ApJ, 787, 16
Pedlar, A., Muxlow, T. W. B., Garrett, M. A., et al. 1999, MNRAS, 307, 761
Peng, Y.-j., Lilly, S. J., Kovac?, K., et al. 2010, ApJ, 721, 193
Pereira-Santaella, M., Colina, L., Garc??a-Burillo, S., et al. 2016, A&A, 594, A81
Pilbratt, G. L., Riedinger, J. R., Passvogel, T., et al. 2010, A&A, 518, 7. http:
//arxiv.org/abs/1005.5331
Poglitsch, A., Waelkens, C., Geis, N., et al. 2010, A&A, 518, L2. http://arxiv.
org/abs/1005.1487
Pound, M. W., & Wolfire, M. G. 2008, in APSC, Vol. 394, Astronomical Data
Analysis Software and Systems XVII, ed. R. W. Argyle, P. S. Bunclark, & J. R.
Lewis, 654
217
Puccetti, S., Comastri, A., Fiore, F., et al. 2014, ApJ, 793, 26
Quillen, A. C., Brookes, M. H., Keene, J., et al. 2006, ApJ, 645, 1092
Radomski, J. T., Packham, C., Levenson, N. A., et al. 2008, ApJ, 681, 141
Renzini, A., & Peng, Y.-j. 2015, ApJ, 801, L29
Richings, A. J., & Faucher-Giguere, C.-A. 2017, ArXiv e-prints, arXiv:1706.03784
Roche, P. F., Whitmore, B., Aitken, D. K., & Phillips, M. M. 1984, MNRAS, 207,
35
Rogstad, D. H., Lockhart, I. A., & Wright, M. C. H. 1974, ApJ, 193, 309
Roussel, H., Wilson, C. D., Vigroux, L., et al. 2010, A&A, 518, L66
Rubin, K. H. R., Prochaska, J. X., Koo, D. C., et al. 2014, ApJ, 794, 156
Rubin, K. H. R., Weiner, B. J., Koo, D. C., et al. 2010, ApJ, 719, 1503
Rubin, R. H., Dufour, R. J., & Walter, D. K. 1993, ApJ, 413, 242
Rubin, R. H., Simpson, J. P., Erickson, E. F., & Haas, M. R. 1988, ApJ, 327, 377
Rupke, D. 2018, Galaxies, 6, 138
Rupke, D., Gu?ltekin, K., & Veilleux, S. 2017, ArXiv e-prints, arXiv:1708.05139
Rupke, D. S., Veilleux, S., & Sanders, D. B. 2002, ApJ, 570, 588
?. 2005a, ApJS, 160, 115
?. 2005b, ApJS, 160, 87
?. 2005c, ApJ, 632, 751
Rupke, D. S. N., & Veilleux, S. 2011, ApJ, 729, L27
?. 2013, ApJ, 768, 75
Rupke, D. S. N., & Veilleux, S. 2015, ApJ, 801, 126
Sanders, D. B., & Mirabel, I. F. 1996, ARA&A, 34, 749
Savage, B. D., & Sembach, K. R. 1996, ARA&A, 34, 279
Scheuer, P. A. G. 1974, MNRAS, 166, 513
Schiano, A. V. R. 1985, ApJ, 299, 24
Schinnerer, E., Eckart, A., Tacconi, L. J., Genzel, R., & Downes, D. 2000, ApJ, 533,
850
218
Scho?ier, F. L., van der Tak, F. F. S., van Dishoeck, E. F., & Black, J. H. 2005,
A&A, 432, 369
Schreier, E. J., Burns, J. O., & Feigelson, E. D. 1981, ApJ, 251, 523
Schurch, N. J., Roberts, T. P., & Warwick, R. S. 2002, MNRAS, 335, 241
Schwartz, C. M., & Martin, C. L. 2004, ApJ, 610, 201
Scoville, N. Z., Soifer, B. T., Neugebauer, G., et al. 1985, ApJ, 289, 129
Sebastian, B., Kharb, P., O?Dea, C. P., Colbert, E. J. M., & Baum, S. A. 2019, ApJ,
883, 189
Sharp, R. G., & Bland-Hawthorn, J. 2010, ApJ, 711, 818
Shimizu, T. T., Mele?ndez, M., Mushotzky, R. F., et al. 2016, MNRAS, 456, 3335
Shimizu, T. T., Mushotzky, R. F., Mele?ndez, M., Koss, M., & Rosario, D. J. 2015,
MNRAS, 452, 1841
Shopbell, P. L., & Bland-Hawthorn, J. 1998, ApJ, 493, 129
Skinner, C. J., Smith, H. A., Sturm, E., et al. 1997, Nature, 386, 472
Smajic?, S., Fischer, S., Zuther, J., & Eckart, A. 2012, A&A, 544, A105
Sofia, U. J., Lauroesch, J. T., Meyer, D. M., & Cartledge, S. I. B. 2004, ApJ, 605,
272
Solomon, P. M., Downes, D., & Radford, S. J. E. 1992, ApJ, 398, L29
Somerville, R. S., & Dave?, R. 2015, ARA&A, 53, 51
Somerville, R. S., Hopkins, P. F., Cox, T. J., Robertson, B. E., & Hernquist, L.
2008, MNRAS, 391, 481
Spinoglio, L., & Malkan, M. A. 1992, ApJ, 399, 504
Spinoglio, L., Malkan, M. A., Smith, H. A., Gonza?lez-Alfonso, E., & Fischer, J.
2005, ApJ, 623, 123
Spinoglio, L., Pereira-Santaella, M., Dasyra, K. M., et al. 2015, ApJ, 799, 21
Spoon, H. W. W., Armus, L., Marshall, J. A., et al. 2009, ApJ, 693, 1223
Spoon, H. W. W., & Holt, J. 2009, ApJ, 702, L42
Spoon, H. W. W., Marshall, J. A., Houck, J. R., et al. 2007, ApJ, 654, L49
Spoon, H. W. W., Farrah, D., Lebouteiller, V., et al. 2013, ApJ, 775, 127
219
Springel, V. 2005, MNRAS, 364, 1105
Stasin?ska, G., Izotov, Y., Morisset, C., & Guseva, N. 2015, A&A, 576, A83
Sternberg, A., & Dalgarno, A. 1995, ApJS, 99, 565
Stone, M., Veilleux, S., Mele?ndez, M., et al. 2016, ApJ, 826, 111
Storey, J. W. V., Watson, D. M., & Townes, C. H. 1981, ApJ, 244, L27
Strateva, I., Ivezic?, Z?., Knapp, G. R., et al. 2001, AJ, 122, 1861
Strickland, D. K., & Heckman, T. M. 2007, ApJ, 658, 258
Strickland, D. K., Heckman, T. M., Weaver, K. A., Hoopes, C. G., & Dahlem, M.
2002, ApJ, 568, 689
Strickland, D. K., & Stevens, I. R. 2000, MNRAS, 314, 511
Sturm, E., Rupke, D., Contursi, A., et al. 2006, ApJ, 653, L13
Sturm, E., Gonza?lez-Alfonso, E., Veilleux, S., et al. 2011, ApJ, 733, L16
Swinyard, B., Nakagawa, T., Merken, P., et al. 2009, Experimental Astronomy, 23,
193
Tadhunter, C., Morganti, R., Rose, M., Oonk, J. B. R., & Oosterloo, T. 2014,
Nature, 511, 440
Tielens, A. G. G. M., & Hollenbach, D. 1985, ApJ, 291, 722
Tingay, S. J., Jauncey, D. L., Reynolds, J. E., et al. 1998, AJ, 115, 960
Tommasin, S., Spinoglio, L., Malkan, M. A., & Fazio, G. 2010, ApJ, 709, 1257
Tremonti, C. A., Moustakas, J., & Diamond-Stanic, A. M. 2007, ApJ, 663, L77
Trotter, A. S., Greenhill, L. J., Moran, J. M., et al. 1998, ApJ, 495, 740
Tully, R. B., Rizzi, L., Shaya, E. J., et al. 2009, AJ, 138, 323
Turner, B. E. 1985, ApJ, 299, 312
Turner, T. J., & Pounds, K. A. 1989, MNRAS, 240, 833
Vasudevan, R. V., & Fabian, A. C. 2007, MNRAS, 381, 1235
?. 2009, MNRAS, 392, 1124
Vasudevan, R. V., Fabian, A. C., Gandhi, P., Winter, L. M., & Mushotzky, R. F.
2010, MNRAS, 402, 1081
Veilleux, S. 1991a, ApJS, 75, 357
220
?. 1991b, ApJS, 75, 383
?. 1991c, ApJ, 369, 331
Veilleux, S. 2006, in Astronomical Society of the Pacific Conference Series, Vol. 357,
Astronomical Society of the Pacific Conference Series, ed. L. Armus & W. T.
Reach, 231
Veilleux, S., & Bland-Hawthorn, J. 1997, ApJ, 479, L105
Veilleux, S., Bolatto, A., Tombesi, F., et al. 2017, ApJ, 843, 18
Veilleux, S., Cecil, G., & Bland-Hawthorn, J. 2005, ARA&A, 43, 769
Veilleux, S., Cecil, G., Bland-Hawthorn, J., et al. 1994, ApJ, 433, 48
Veilleux, S., Kim, D.-C., & Sanders, D. B. 1999, ApJ, 522, 113
Veilleux, S., Kim, D. C., & Sanders, D. B. 2002, ApJS, 143, 42. http://arxiv.
org/abs/astro-ph/0207401
Veilleux, S., Maiolino, R., Bolatto, A. D., & Aalto, S. 2020, arXiv e-prints,
arXiv:2002.07765
Veilleux, S., Shopbell, P. L., Rupke, D. S., Bland -Hawthorn, J., & Cecil, G. 2003,
AJ, 126, 2185
Veilleux, S., Rupke, D. S. N., Kim, D.-C., et al. 2009, ApJS, 182, 628
Veilleux, S., Mele?ndez, M., Sturm, E., et al. 2013, ApJ, 776, 27
Venturi, G., Marconi, A., Mingozzi, M., et al. 2017, Frontiers in Astronomy and
Space Sciences, 4, 46
Ve?ron-Cetty, M.-P., & Ve?ron, P. 2006, A&A, 455, 773
Voit, G. M. 1992, ApJ, 399, 495
Walter, F., Weiss, A., & Scoville, N. 2002, ApJ, 580, L21
Walter, F., Bolatto, A. D., Leroy, A. K., et al. 2017, ApJ, 835, 265
Weiner, B. J., Coil, A. L., Prochaska, J. X., et al. 2009, ApJ, 692, 187
Werner, M. W., Roellig, T. L., Low, F. J., et al. 2004, ApJS, 154, 1
Westmoquette, M. S., Smith, L. J., Gallagher, J. S., & Walter, F. 2013, MNRAS,
428, 1743
Westmoquette, M. S., Smith, L. J., & Gallagher, III, J. S. 2011, MNRAS, 414, 3719
Whitaker, K. E., Franx, M., Leja, J., et al. 2014, ApJ, 795, 104
221
Wills, K. A., Redman, M. P., Muxlow, T. W. B., & Pedlar, A. 1999, MNRAS, 309,
395
Wilson, A. S., & Ulvestad, J. S. 1987, ApJ, 319, 105
Wilson, C. D., Petitpas, G. R., Iono, D., et al. 2008, ApJS, 178, 189
Winter, L. M., Veilleux, S., McKernan, B., & Kallman, T. R. 2012, ApJ, 745, 107
Wolfire, M. G., Tielens, A. G. G. M., & Hollenbach, D. 1990, ApJ, 358, 116
Wright, E. L. 2006, PASP, 118, 1711
Yamauchi, A., Nakai, N., Sato, N., & Diamond, P. 2004, PASJ, 56, 605
Zakamska, N. L., Hamann, F., Pa?ris, I., et al. 2016, MNRAS, 459, 3144
Zhang, D. 2018, Galaxies, 6, 114
Zschaechner, L. K., Walter, F., Bolatto, A., et al. 2016, ApJ, 832, 142
Zschaechner, L. K., Bolatto, A. D., Walter, F., et al. 2018, ApJ, 867, 111
Zubovas, K., & Nayakshin, S. 2014, MNRAS, 440, 2625
222