1. Gas Disks and Supermassive Black Holes in Nearby Radio Galaxies
Jacob Noel-Storr
Advisors Jacqueline H. van Gorkom
and Stefi A. Baum
Submitted in partial fulfillment of the
requirements for the degree
of Doctor of Philosophy
in the Graduate School of Arts and Sciences.
COLUMBIA UNIVERSITY
2004

2. c 2004
Jacob Noel-Storr
All Rights Reserved

3. Abstract
Gas Disks and Supermassive Black Holes in Nearby Radio Galaxies
Jacob Noel-Storr
We present a detailed analysis of a set of medium resolution spectra, obtained by
the Space Telescope Imaging Spectrograph on board the Hubble Space Telescope, of
the emission line gas present in the nuclei of a complete sample of 21 nearby, early-type
galaxies with radio jets (the UGC FR-I sample). For each galaxy nucleus we present
spectroscopic data in the region of Hydrogen-alpha and the derived kinematics.
We find that is 67% of the nuclei the gas appears to be rotating and, with one
exception, the cases where rotation is not seen are either face on or have complex
central morphologies. We find that in 62% of the nuclei the fit to the central spectrum
is improved by the inclusion of a broad component. The broad components have a
mean velocity dispersion of 1349 ± 345 km s−1
and are redshifted from the narrow
line components (assuming an origin in Hydrogen-alpha) by 486 ± 443 km s−1
.
We generated model velocity profiles for the nuclei including no black hole, a
1 × 108
solar mass black hole and a 9 × 108
solar mass black hole. We compared the

4. predicted profiles to the observed velocity profiles from the above spectra and found
kinematic signatures compatible with black holes > 1 × 108
solar masses in 53% of
the sample (in the remaining galaxies we are unable to rule in or out the presence of
a nuclear black hole). We suspect that flow is a significant factor in the nucleus of
NGC 2329. We found hints of jet-disk interaction in 24% of the sample nuclei and
signs of twists or warps in 19% of the sample nuclei. 24% of the velocity profiles show
signs of multiple distinct kinematic components. We suggest that the gas disks in
nearby radio galaxies are generally not well settled systems in the equitorial plane of
the gravitational potentials.
We characterize the kinematic state of the nuclear gas through three weighted
mean parameters, and find that again the disks appear not to be well settled. We
show evidence of a connection between the stellar and gas velocity dispersions. We
show correlations in the nuclear fluxes from the Radio to the X-ray regimes, suggesting
a common origin (e.g. accretion disks or jet bases) for the nuclear fluxes. We find
agreement with the work of others that the nuclear flux and black hole masses are not
correlated, though the width of the broad components, described above may correlate
with the nuclear flux.

5. Contents
1 Introduction 1
1.1 Radio galaxies 3
1.2 Nuclear dust and gas 6
1.3 Connections to supermassive black holes 10
1.4 This dissertation 14
2 The UGC FR-I Sample 26
2.1 Sample selection and properties 26
2.2 Multiwavelength observations 28
2.2.1 Radio properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.2.2 Optical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3 Stellar Dynamics 30
3 STIS spectroscopy of the emission line gas in the nuclei
i

6. of nearby FR-I galaxies 57
3.1 Introduction 57
3.2 STIS Observations 58
3.2.1 Data Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.3 Analysis 62
3.3.1 Single Gaussian line fitting . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3.2 Fits with an additional free component . . . . . . . . . . . . . . . . . 65
3.3.3 Quantifying error sources . . . . . . . . . . . . . . . . . . . . . . . . 67
3.4 Individual galaxy descriptions 69
3.5 Interpretation and Discussion 79
3.5.1 Rotators and non-rotators . . . . . . . . . . . . . . . . . . . . . . . . 79
3.5.2 Flux ratios and ionization . . . . . . . . . . . . . . . . . . . . . . . . 83
3.5.3 Are we observing broad lines? . . . . . . . . . . . . . . . . . . . . . . 83
3.5.4 Constraining the line shapes . . . . . . . . . . . . . . . . . . . . . . . 85
3.6 Conclusions 87
4 Modeling gas in gravitational potentials 147
4.1 Introduction 147
4.2 Thin disk models with and without a black hole 149
4.2.1 WFPC2 imaging: stars and dust . . . . . . . . . . . . . . . . . . . . . 150
4.2.2 Flux distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
ii

7. 4.2.3 Stellar luminosity densities and mass distributions . . . . . . . . . . . 152
4.2.4 Dynamical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
4.2.5 Locating the zero point velocities . . . . . . . . . . . . . . . . . . . . 157
4.2.6 Sensitivity to parameters . . . . . . . . . . . . . . . . . . . . . . . . 158
4.3 Descriptions of individual galaxies 161
4.4 Discussion 172
4.4.1 Description of the central kinematics . . . . . . . . . . . . . . . . . . 172
4.4.2 Settling of the gas into a thin disk . . . . . . . . . . . . . . . . . . . . 174
4.4.3 Drivers of unsettled motion . . . . . . . . . . . . . . . . . . . . . . . 176
4.5 Conclusions 179
5 Nuclear gas kinematics and central engines 234
5.1 Introduction 234
5.2 Global kinematic parameterizations 236
5.2.1 Effects of inclination . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
5.2.2 Correlations between kinematic parameters . . . . . . . . . . . . . . . 242
5.2.3 Relationships to the stellar kinematics . . . . . . . . . . . . . . . . . 244
5.3 The central engines 246
5.3.1 Are the Broad Lines physical? . . . . . . . . . . . . . . . . . . . . . . 247
5.3.2 Correlation of nuclear fluxes . . . . . . . . . . . . . . . . . . . . . . . 249
5.3.3 Correlations between host galaxy and central engine . . . . . . . . . . 250
5.4 Conclusions 252
iii

8. 6 Conclusions 272
6.1 Summary 272
6.2 Future directions 276
References 284
iv

9. List of Tables
1.1 Properties of various classes of Active Galactic Nuclei. . . . . . . . . 16
1.2 Reliable black hole mass measurements. . . . . . . . . . . . . . . . . . 17
2.1 Properties of the galaxy sample members. . . . . . . . . . . . . . . . 32
2.2 Multiwavelength Fluxes of the UGC FR-I Sample Galaxies. . . . . . . 33
2.3 Stellar velocity dispersions and estimated black hole masses. . . . . . 34
3.1 HST-STIS G750M observing log for this program. . . . . . . . . . . . 89
3.2 HST/STIS Instrumental properties for the configurations used. . . . 90
3.3 Position angles of various axes. . . . . . . . . . . . . . . . . . . . . . 91
3.4 Spectral Lines in the region of Hα. . . . . . . . . . . . . . . . . . . . 92
3.5 NGC 193: Measured Parameters. . . . . . . . . . . . . . . . . . . . . 93
3.6 NGC 315: Measured Parameters. . . . . . . . . . . . . . . . . . . . . 94
3.7 NGC 383: Measured Parameters. . . . . . . . . . . . . . . . . . . . . 95
3.8 NGC 541: Measured Parameters. . . . . . . . . . . . . . . . . . . . . 96
3.9 NGC 741: Measured Parameters. . . . . . . . . . . . . . . . . . . . . 97
3.10 UGC 01841: Measured Parameters. . . . . . . . . . . . . . . . . . . 98
3.11 NGC 2329: Measured Parameters. . . . . . . . . . . . . . . . . . . . 99
3.12 NGC 2892: Measured Parameters. . . . . . . . . . . . . . . . . . . . 100
3.13 NGC 3801: Measured Parameters. . . . . . . . . . . . . . . . . . . . 101
v

10. 3.14 NGC 3862: Measured Parameters. . . . . . . . . . . . . . . . . . . . 102
3.15 UGC 7115: Measured Parameters. . . . . . . . . . . . . . . . . . . . 103
3.16 NGC 4261: Measured Parameters. . . . . . . . . . . . . . . . . . . . 104
3.17 NGC 4335: Measured Parameters. . . . . . . . . . . . . . . . . . . . 105
3.18 M84 : Measured Parameters. . . . . . . . . . . . . . . . . . . . . . . 106
3.19 NGC 4486: Measured Parameters. . . . . . . . . . . . . . . . . . . . 107
3.20 NGC 5127: Measured Parameters. . . . . . . . . . . . . . . . . . . . 108
3.21 NGC 5141: Measured Parameters. . . . . . . . . . . . . . . . . . . . 109
3.22 NGC 5490: Measured Parameters. . . . . . . . . . . . . . . . . . . . 110
3.23 NGC 7052: Measured Parameters. . . . . . . . . . . . . . . . . . . . 111
3.24 UGC 12064: Measured Parameters. . . . . . . . . . . . . . . . . . . 112
3.25 NGC 7626: Measured Parameters. . . . . . . . . . . . . . . . . . . . 113
3.26 Effect of making various fit parameters free. . . . . . . . . . . . . . . 114
3.27 Presence of a Nuclear Broad Line. . . . . . . . . . . . . . . . . . . . . 115
3.28 Fits to the central pixel for each galaxy, including broad lines. Kine-
matics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
3.29 Fits to the central pixel for each galaxy, including broad lines. Fluxes. 117
3.30 Kinematic estimators within 100 pc of the nucleus. . . . . . . . . . . 118
3.31 Comparison of broad line statistics . . . . . . . . . . . . . . . . . . . 119
3.32 Kinematic parameters measured using various free-parameter sets. . 120
4.1 Disk inclinations and dust masses. . . . . . . . . . . . . . . . . . . . 182
4.2 STIS PSF Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . 183
4.3 Luminosity density fit parameters. . . . . . . . . . . . . . . . . . . . 184
4.4 Mass to light ratio (Υ) estimates for sample galaxies. . . . . . . . . . 185
4.5 Best fitting velocity offsets. . . . . . . . . . . . . . . . . . . . . . . . 186
4.6 Black hole signatures in the central kinematics. . . . . . . . . . . . . 187
vi

11. 5.1 Weighted mean kinematic parameters. . . . . . . . . . . . . . . . . . 255
5.2 Correlation of X-ray flux with other nuclear flux parameters. . . . . 256
5.3 Correlations between kinematic parameters and nuclear fluxes. . . . . 257
vii

12. List of Figures
1.1 A cartoon radio galaxy. . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.2 An example of an FR-I and an FR-II radio galaxy . . . . . . . . . . . 19
1.3 The Unified Scheme for Active Galactic Nuclei. . . . . . . . . . . . . 20
1.4 Non uniform dust disks in NGC 383. . . . . . . . . . . . . . . . . . . 21
1.5 The segregation of AGNs based on the circumnuclear gas disk. . . . . 22
1.6 Feeding the central engine. . . . . . . . . . . . . . . . . . . . . . . . . 23
1.7 Velocity cusp in the nucleus of M87. . . . . . . . . . . . . . . . . . . . 24
1.8 The M• − σc relation. . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1 Optical continuum image of the nuclear region of NGC 193. . . . . . 35
2.2 Optical continuum image of the nuclear region of NGC 315. . . . . . 36
2.3 Optical continuum image of the nuclear region of NGC 383. . . . . . 37
2.4 Optical continuum image of the nuclear region of NGC 541. . . . . . 38
2.5 Optical continuum image of the nuclear region of NGC 741. . . . . . 39
2.6 Optical continuum image of the nuclear region of UGC 1841. . . . . . 40
2.7 Optical continuum image of the nuclear region of NGC 2329. . . . . . 41
2.8 Optical continuum image of the nuclear region of NGC 2892. . . . . . 42
2.9 Optical continuum image of the nuclear region of NGC 3801. . . . . . 43
2.10 Optical continuum image of the nuclear region of NGC 3862. . . . . . 44
2.11 Optical continuum image of the nuclear region of UGC 7115. . . . . . 45
viii

13. 2.12 Optical continuum image of the nuclear region of NGC 4261. . . . . . 46
2.13 Optical continuum image of the nuclear region of NGC 4335. . . . . . 47
2.14 Optical continuum image of the nuclear region of M84. . . . . . . . . 48
2.15 Optical continuum image of the nuclear region of NGC 4486. . . . . . 49
2.16 Optical continuum image of the nuclear region of NGC 5127. . . . . . 50
2.17 Optical continuum image of the nuclear region of NGC 5141. . . . . . 51
2.18 Optical continuum image of the nuclear region of NGC 5490. . . . . . 52
2.19 Optical continuum image of the nuclear region of NGC 7052. . . . . . 53
2.20 Optical continuum image of the nuclear region of UGC 12064. . . . . 54
2.21 Optical continuum image of the nuclear region of NGC 7626. . . . . . 55
2.22 The M• − σc relation. . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.1 Key to observation and fit data plots. . . . . . . . . . . . . . . . . . . 121
3.2 Observation and fit data for NGC 193 . . . . . . . . . . . . . . . . . 122
3.3 Observation and fit data for NGC 315 . . . . . . . . . . . . . . . . . 123
3.4 Observation and fit data for NGC 383 . . . . . . . . . . . . . . . . . 124
3.5 Observation and fit data for NGC 541 . . . . . . . . . . . . . . . . . 125
3.6 Observation and fit data for NGC 741 . . . . . . . . . . . . . . . . . 126
3.7 Observation and fit data for UGC 1841 . . . . . . . . . . . . . . . . . 127
3.8 Observation and fit data for NGC 2329 . . . . . . . . . . . . . . . . . 128
3.9 Observation and fit data for NGC 2892 . . . . . . . . . . . . . . . . . 129
3.10 Observation and fit data for NGC 3801 . . . . . . . . . . . . . . . . . 130
3.11 Observation and fit data for NGC 3862 . . . . . . . . . . . . . . . . . 131
3.12 Observation and fit data for UGC 7115 . . . . . . . . . . . . . . . . . 132
3.13 Observation and fit data for NGC 4261 . . . . . . . . . . . . . . . . . 133
3.14 Observation and fit data for NGC 4335 . . . . . . . . . . . . . . . . . 134
ix

14. 3.15 Observation and fit data for NGC 4374 . . . . . . . . . . . . . . . . . 135
3.16 Observation and fit data for NGC 4486 . . . . . . . . . . . . . . . . . 136
3.17 Observation and fit data for NGC 5127 . . . . . . . . . . . . . . . . . 137
3.18 Observation and fit data for NGC 5141 . . . . . . . . . . . . . . . . . 138
3.19 Observation and fit data for NGC 5490 . . . . . . . . . . . . . . . . . 139
3.20 Observation and fit data for NGC 7052 . . . . . . . . . . . . . . . . . 140
3.21 Observation and fit data for UGC 12064 . . . . . . . . . . . . . . . . 141
3.22 Observation and fit data for NGC 7626 . . . . . . . . . . . . . . . . . 142
3.23 Difference in mean velocity within 100 pc of each side of the nucleus . 143
3.24 Difference in mean velocity within 100 pc of each side of the nucleus . 144
3.25 Mean gas velocity dispersion within 100 pc of the nucleus . . . . . . . 145
3.26 [N II] against Hα fluxes for the UGC FR-I sample members . . . . . . 146
4.1 Data-Model residuals with varying velocity offsets for NGC 193. . . . 188
4.2 Data-Model residuals with varying velocity offsets for NGC 315. . . . 189
4.3 Data-Model residuals with varying velocity offsets for NGC 383. . . . 190
4.4 Data-Model residuals with varying velocity offsets for NGC 541. . . . 191
4.5 Data-Model residuals with varying velocity offsets for NGC 741. . . . 192
4.6 Data-Model residuals with varying velocity offsets for UGC 1841. . . 193
4.7 Data-Model residuals with varying velocity offsets for NGC 2329. . . 194
4.8 Data-Model residuals with varying velocity offsets for NGC 2892. . . 195
4.9 Data-Model residuals with varying velocity offsets for NGC 3862. . . 196
4.10 Data-Model residuals with varying velocity offsets for NGC 4335. . . 197
4.11 Data-Model residuals with varying velocity offsets for M84. . . . . . . 198
4.12 Data-Model residuals with varying velocity offsets for NGC 4486. . . 199
4.13 Data-Model residuals with varying velocity offsets for NGC 5127. . . 200
x

15. 4.14 Data-Model residuals with varying velocity offsets for NGC 5141. . . 201
4.15 Data-Model residuals with varying velocity offsets for NGC 5490. . . 202
4.16 Data-Model residuals with varying velocity offsets for NGC 7052. . . 203
4.17 Data-Model residuals with varying velocity offsets for UGC 12064. . . 204
4.18 Data-Model residuals with varying velocity offsets for NGC 7626. . . 205
4.19 Effect of changing the STIS PSF. . . . . . . . . . . . . . . . . . . . . 206
4.20 Central observations of M87. . . . . . . . . . . . . . . . . . . . . . . . 207
4.21 Varying Υ, q and i in models of NGC 193. . . . . . . . . . . . . . . . 208
4.22 Varying Υ, q and i in models of NGC 4335. . . . . . . . . . . . . . . 209
4.23 Observed and modeled velocity profiles for NGC 193. . . . . . . . . . 210
4.24 Observed and modeled velocity profiles for NGC 315. . . . . . . . . . 211
4.25 Observed and modeled velocity profiles for NGC 383. . . . . . . . . . 212
4.26 Observed and modeled velocity profiles for NGC 541. . . . . . . . . . 213
4.27 Observed and modeled velocity profiles for NGC 741. . . . . . . . . . 214
4.28 Observed and modeled velocity profiles for UGC 1841. . . . . . . . . 215
4.29 Observed and modeled velocity profiles for NGC 2329. . . . . . . . . 216
4.30 Observed and modeled velocity profiles for NGC 2892. . . . . . . . . 217
4.31 Observed and modeled velocity profiles for NGC 3862. . . . . . . . . 218
4.32 Observed and modeled velocity profiles for NGC 4261. . . . . . . . . 219
4.33 Observed and modeled velocity profiles for NGC 4335. . . . . . . . . 220
4.34 Observed and modeled velocity profiles for M84. . . . . . . . . . . . . 221
4.35 Observed and modeled velocity profiles for NGC 4486. . . . . . . . . 222
4.36 Observed and modeled velocity profiles for NGC 5127. . . . . . . . . 223
4.37 Observed and modeled velocity profiles for NGC 5141. . . . . . . . . 224
4.38 Observed and modeled velocity profiles for NGC 5490. . . . . . . . . 225
4.39 Observed and modeled velocity profiles for NGC 7052. . . . . . . . . 226
xi

16. 4.40 Observed and modeled velocity profiles for UGC 12064. . . . . . . . . 227
4.41 Observed and modeled velocity profiles for NGC 7626. . . . . . . . . 228
4.42 NGC 4335, observations with jet locations indicated. . . . . . . . . . 229
4.43 NGC 7626, observations with jet locations indicated. . . . . . . . . . 230
4.44 NGC 193, observations with jet locations indicated. . . . . . . . . . . 231
4.45 UGC 12064, observations with jet locations indicated. . . . . . . . . . 232
4.46 The M• − σc relation with UGC FR-I black hole limits indicated. . . 233
5.1 Difference in weighted mean velocity within 100 pc on each side of the
nucleus in the central slit as a function of dust axis ratio. . . . . . . . 258
5.2 Weighted mean gas velocity dispersion along the central slit of each
galaxy as a function of dust axis ratio. . . . . . . . . . . . . . . . . . 259
5.3 Weighted mean point-to-point variations in gas velocity along the cen-
tral slit as a function of dust disk axis ratio. . . . . . . . . . . . . . . 260
5.4 σ100
2
/∆2
100 as a function of dust axis ratio. . . . . . . . . . . . . . . . 261
5.5 σ100 as a function of ∆100 for the sample nuclei. . . . . . . . . . . . . 262
5.6 100 as a function of ∆100 for the sample nuclei. . . . . . . . . . . . . 263
5.7 100 as a function of σ100 for the sample nuclei. . . . . . . . . . . . . . 264
5.8 ∆100 for each nucleus as a function of the central stellar velocity dis-
persion (σc). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
5.9 σ100 for each nucleus as a function of the central stellar velocity dis-
persion (σc). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
5.10 100 for each nucleus as a function of the central stellar velocity disper-
sion (σc). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
5.11 σ100
2
/∆2
100 for each nucleus as a function of the central stellar velocity
dispersion (σc). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
5.12 Correlations between nuclear fluxes. . . . . . . . . . . . . . . . . . . . 269
5.13 Nuclear SEDs for 5 of the UGC FR-I sample galaxies. . . . . . . . . . 270
5.14 Model SEDs for M87. . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
xii

17. Acknowledgments
First I would like, of course, to thank my parents, Jenny and Peter, and my entire
family. I must also thank my scientific collaborators, without whom this thesis would
not have been even remotely possible: Gijs Verdoes Kleijn, Stefi Baum, Jacqueline
van Gorkom, Chris O’Dea, Roeland van der Marel, Tim de Zeeuw, and Marcella
Carollo. I also thank many wonderful astronomers for discussions on these topics -
in particular Aaron Barth and Dave Axon, and also Johann Knapen, Luis Ho and
Thaisa Storchi-Bergmann for organizing fantastic conferences that have taken me
from start to finish of this thesis.
Everyone that I have known in the astronomy department, you have been fantastic.
In particular thanks to Aeree Chung and Stephen Muchovej, your friendship is very
important to me... and we will always have Rio! I also want to give particular thanks
to Millie Kramer-Garcia for being possibly the most outstanding, amazing, awesome
department administrator in the history of the entire universe. Probably without you
none of us graduate students would make it past one week of grad school.
Mark Wainer, Simon Stam, Daniel Baker and Alex Vitti – thank you guys so
much for always keeping me going, and for keeping me smiling. Your friendship has
been outstanding, and I will value it forever.
I want to thank fantastic friends that I have had through the years – Ethan
Hurdus, Neil Corbett, Jon Doyle, Stuart Laverack, Eileen Brain, Jackie Bletcher,
Jonathan Bletcher, Adam Pogash, Daniel Pogash, Carrie Johnston, Lucy Edge, Abi
Bralee, Tzu-Ching Chang, Andy Jacobowitz, Karen Vanlandingham, Ben Clissold,
Solomon Meltser, Gina Brissenden, and John Drury – Just for being amazing people.
Gareth Repton, Jon Soyt, Matt Schoen, Steven Turner, Adam Simon, Daniel
Bryer, Eric Lunin, Ariel Stukalin, Cliff Shapiro, Seth Kahn, Scott Menke, Ben Sil-
verman, James Dawson, Isaac Gerstein, Peter Plant, Freddie Claro, Stephen Con-
lon, Brian Bolin, Alex Chval, Andrew Prouty, T.J. Wagner, Jake Skinner, Michael
Seiler, Billie Swift, Heather Groch, Dan Balick, Tim Mendenko, Jeff Wincek, Charles
Stanton-Jones, Eilat Glikman, Suvi Gezari, Andreea Petric, Jen Donovan, Mark Di-
jkstra, Sarah Tuttle, Ben Johnson, and Antara Basu-Zych – You guys rule! (James
Dawson, you have been the man through the last few weeks of getting this dissertation
written - Thanks dude).
Everyone else I didn’t have space to write about – Thank you all too!!
Thank you.
xiii

18. To:
Lofty
Jenny, Peter
Everyone who has ever set foot at Camp Watonka and made it feel like home year
after year, and especially all of my campers.
and Spencer and Ceci... In 21 years I expect to be reading your PhD theses!
xiv

19. Chapter 1
Introduction
The relationship between quiescent and active early type galaxies bears on our un-
derstanding of black holes and their role in the active nuclei of galaxies, and on the
evolution of galactic nuclei along with their host galaxies. Nearby early-type galaxies
with radio jets provide an opportunity to gain an understanding of the conditions in
a galaxy which lead to the formation of a radio-active nucleus and of the physics of
the regions which harbor the black hole and jet-formation regimes.
In this dissertation, we present a detailed analysis of the kinematics of the emission
line gas in the nuclei of a complete sample of nearby, radio galaxies. This forms a
part of a larger, coordinated, multi-wavelength study of the same sample, which is
described in the next chapter.
1

20. 2
The study of radio galaxies addresses some important astrophysical questions that
pertain to the nature of active galaxies in general:
• How does fuel get into the central engine? How do active and inactive galaxies
differ in terms of fuel and fueling?
• What connections exist between host galaxies, central engines and central su-
permassive black holes?
• How do radio galaxies fit into ‘unified schemes’ that describe various types of
active galaxy?
In this chapter we will first introduce the reader to radio galaxies, which are strong
radio sources associated with luminous elliptical galaxies, next our discussion moves
inwards to focus on the gas and dust that has been found in many of radio galaxy
nuclei. We increase magnification still further as we go on to talk about the search
for supermassive black holes and the connections they have to the galaxy as a whole,
particularly relationships between the black hole masses (M•) and the large scale
stellar kinematics. We close by outlining the contents of this dissertation.

21. 3
1.1 Radio galaxies
A radio galaxy can produce a jet-lobe, radio wavelength emitting, structure on mega-
parsec scales, emerging from a central engine that resides in an active region no more
than a few milli-parsecs in size. The radio emission is well understood to be from
synchrotron radiation, but the processes involved in fueling and collimating the ra-
dio jets are yet to be explained (Lynden-Bell, 2001). The energy that powers the
radio jets is typically believed to be produced during the accretion of material onto
a central supermassive black hole (e.g., Lynden-Bell, 1969). The formation of such
well organized large scale jet structures is some of the most compelling evidence for
the necessity of supermassive black holes as an ingredient in the makeup of a radio
galaxy (Rees, 1984).
Radio emission can be detected even from relatively quiescent galaxies, such
as the Milky Way, on the level of around 1037
erg s−1
. Active galaxies (such as
Seyferts or starburst galaxies) radiate at around 1037
erg s−1
, and radio galaxies (and
also quasars) extend this energy output range to beyond 1045
erg s−1
(∼ 1012
L )
(Burke & Graham-Smith, 1997). An active galaxy central engine capable of releas-
ing ∼ 0.1Mfuelc2
of infalling fuel as radiation would need to be fueled at a rate of
∼> 1 M yr−1
to maintain this luminosity.

22. 4
The ‘cartoon view’ of a typical radio galaxy structure contains a bright core from
which emanate two jets of radio emitting material which terminate in two large radio
lobes. (Figure 1.1 shows such a cartoon view). Two classes of radio galaxy were
defined by Fanaroff & Riley (1974) based on the morphology of the radio emission.
Fanaroff-Riley Class I galaxies (FR-Is) are brightest at the inner jets, with emission
that becomes gradually fainter out into the radio lobes, while in the more powerful
FR-II galaxies, hot spots at the outer edges of the radio lobes dominate the emission.
Owen & Ledlow (1997) present observations of many radio galaxies, illustrating of
course that there are many inadequacies in such a simplified viewpoint; in Figure 1.2
we present examples of a typical FR-I and FR-II galaxy from their sample.
Radio galaxies belong to the class of galaxies that have Active Galactic Nuclei
(AGN) which are characterized by large energy outputs from very small regions in
their cores. Commonly, the ‘Unified Scheme’ (Urry & Padovani, 1995) is used to
explain different properties of the different types of AGN and the connections between
them, principally by observing the model from varying orientations. A representation
of an AGN according to the unified scheme is shown in Figure 1.3. Various types of
AGN and their properties are listed in Table 1.1. The type of AGN and host galaxy
are also linked, radio galaxies being almost exclusively found in early-type (elliptical
or S0) galaxies.

23. 5
This thesis focuses on a sample of nearby radio galaxies (see Chapter 2). The
sample is complete out to redshifts of 7000 km s−1
(100 Mpc) in the northern hemi-
sphere, and contains only FR-I galaxies, as no FR-IIs are found in this volume (there
is only one FR-II within 7000 km s−1
, Centaurus A, which is in the southern sky at a
declination of ∼ −43◦
). Radio galaxies and BL-Lacs provide a high energy contrast to
Seyfert galaxies and LINERs (Low Ionization Narrow line Emission Region galaxies),
and may be close cousins of the Quasars observed at higher redshifts.
Both FR-I and FR-II galaxies have jets that are well collimated on all scales,
though FR-Is tend to be more prone to having twisted and ‘blobby’ jets while FR-IIs
tend to have straighter more uniform jets (see Bridle, 1984). The accepted paradigm
is that the jets start out at relativistic velocities in both FR-Is and FR-IIs, but that
in FR-Is the jets then decelerate on scales less than ∼ 1 kpc. (Bridle & Perley,
1984, provide a review of the phenomenology of extragalactic jets). The most elegant
demonstration that flow may actually takes place along the jets is presented by Biretta
et al. (1999), who while monitoring the jet of M87 over four years were able to trace
the motion of blobs of emission, providing compelling evidence that some type of flow
does occur along the jets.
Baum et al. (1995) describe various differences in the populations of FR-I and FR-

24. 6
II galaxies. They note that at the same host galaxy magnitude or radio luminosity
FR-IIs can produce and order of magnitude more optical line emission than FR-Is.
This led them to conclude that FR-IIs were predominantly ionized by continuum
sources in the AGN, while in FR-Is the host galaxy may also play an important
role in the ionization of the emission line gas. Following earlier work by Rees et al.
(1982), Baum et al. (1995) also propose that FR-I galaxies may be produced when
the central engine is fed at a lower accretion rate and FR-IIs are produced when the
fuel is being accreted more quickly. Understanding the fueling of the central engines
of nearby radio galaxies, or at least better understanding the kinematic nature of the
fuel sources, may lead to greater understanding of the physical mechanisms driving
this dichotomy.
1.2 Nuclear dust and gas
Though when observed on the sky elliptical galaxies can appear smooth and largely
featureless, the Hubble Space Telescope (HST) was able to resolve dust and gas fea-
tures present in their central regions (Jaffe et al., 1993; Ford et al., 1994). Central
emission line dust and gas disks are detected in about 20% of all giant ellipticals,
and virtually all nearby radio galaxies that harbor kiloparsec scale radio jets (e.g.
van Dokkum & Franx, 1995; Verdoes Kleijn et al., 1999; Capetti et al., 2000; de Koff

25. 7
et al., 2000; Tomita et al., 2000; Tran et al., 2001; Laine et al., 2003). This dust and
gas presumably provides a bulk of the fuel for the central engine.
The kinematics of the disk are not well understood, and important questions
remain regarding the importance of non-circular motions (such as turbulence, inflow,
outflow, winds, etc...). It is also of interest whether these disks are short lived or
long lived, and whether they are providing fuel to the central engine or playing a
role in the collimation of the jets. Disks have been identified that are non-uniform in
continuum light and are surrounded by arcs, filaments, and diffuse absorbing clumps,
suggesting that the dust (and by association the gas) in the cores of these galaxies is
not dynamically settled (for example, Martel et al., 2000, ; Figure 1.4).
Assuming that this gas and dust indeed provides the primary fuel source of the
central engine, a long standing problem is the ‘fueling problem’ – that is how to
remove the large angular momentum of the gas in a disk and get it into the central
engine. This problem is described in a well known cartoon by Phinney (1994) of a
baby being fed by a huge (angular momentum) spoon into its tiny mouth (Figure 1.6).
Also, typical findings are that 60% of quiescent early-type galaxies have detections
of emission line gas (e.g. Philips et al., 1986; Goudfrooij et al., 1994) and about 40%
have nuclear dust (e.g. van Dokkum & Franx, 1995; Tran et al., 2001). Despite the

26. 8
apparent presence of fuel, these galaxies fail to produce radio jets or any significant
nuclear activity.
Unified schemes (see above) suggest that the ionized gas disk not intersect the
photo-ionization cone of the central engine, which would require that the disks be
shock ionized (Doptia et al., 1997). An investigation of the ionization structure of the
disk, and any variations, can therefore reveal properties important to understanding
and confirming the position of FR-I galaxies in such schemes. The nature of the gas
disks is differentiated between the different types of AGN with key parameters being
the gas mass, the black hole mass and the star formation rate. Different AGN can be
plotted as a function of these parameters as represented in Figure 1.5. Radio Galaxies
have large black hole masses, low star formation rates and relatively low gas masses
so would fall into the same region of this plot as occupied by the quasars (QSOs).
Different fueling mechanism may operate to bring the fuel into the center depending
on these parameters (see Wada, 2004, for a recent review).
Asymmetries in galaxy potentials can induce dynamical resonances that perturb
the gas dynamics in the central regions. Hydrodynamical simulations (e.g., Ma-
ciejewski, 2004) show that these resonances, and the resulting perturbations in the
gas motions, can be governed primarily by the central black hole and on scales much

27. 9
larger than the generally expected sphere of influence (see Equation 1.1 below). These
results may imply that the central black hole is able to have kinematic influences far
into the gas disk, and may even be able to regulate flow (and hence the fueling of the
central engine) though these mechanisms.
The nuclear gas and dust may be perturbed by sources unrelated to the gravita-
tional potential of the galaxy, for example interactions with jets, starburst activity
and galactic mergers, each of which could have considerable consequences for the
kinematics of the gas, and the organization of structures in the central regions of
these galaxies.
Sol´orzano-I˜narrea et al. (2001) suggest that the emission line gas may be affected
by jet induced shocks even in sources where the radio emission structures are much
larger in extent than the regions of emission line gas (as is the case for the radio
galaxies described in this thesis). It would take a certain amount of misalignment
between jet and disk for such interactions to be possible, and such large misalignments
between disks and jets may be observed in some radio galaxies (Schmitt et al., 2002;
Verdoes Kleijn & de Zeeuw, 2004).
Starbursts and AGN go hand in hand in many galactic nuclei (Beckman, 2001)
and episodes of star burst activity could induce shocks into the gas disks, having a

28. 10
pronounced effect of the observed kinematics. The gas in the emission line disk is
an obvious source of fuel for any starburst activity that may have taken place, which
would have had dramatic consequences for the disk kinematics the remnants of which
may not be fast to dissipate. Galactic mergers in the evolutionary history of the radio
galaxy would also have disrupted the gas: both by the addition of new gas into the
disk and by tidal disruption of the gas by the merging galaxy.
1.3 Connections to supermassive black holes
Current evidence suggests that all galaxies may have a central supermassive black
hole and that the mass of this black hole is a strong function of the mass and lumi-
nosity of the mass spheroid in which it presently resides (for example, Kormendy &
Richstone, 1995), and an even stronger function of the central stellar velocity disper-
sion (Ferrarese & Merritt, 2000; Gebhardt et al., 2000a; Merritt & Ferrarese, 2001;
Tremaine et al., 2002). It is not clear what drives these relationships or whether they
are the same for active and quiescent galaxies. Since black hole growth and nuclear
activity are causally related, scatter in these relationships can, in principle, put limits
on the frequency and duration of nuclear activity in galaxies.
Four principle methods exist for the direct measurement of black hole masses. The

29. 11
most reliable is, of course, by the measurement of the proper motions and accelerations
of the stars in the closest orbits to the black hole. This has now been achieved in
the nucleus of our own galaxy (Ghez et al., 1998; Melia & Falcke, 2001; Reid et al.,
2003; Ghez et al., 2003), though this will be the only case where we can make these
measurements. Kinematics of water masers in the inner nuclei also provide what
appear to be very robust measurements of the black hole mass (Greenhill & Gwinn,
1997; Herrnstein et al., 1999), though cases with suitable maser organization and
orientation prove hard to find. For larger samples, one must use dynamical modeling
either of the stars or of the gas in the nuclear region of the galaxy.
A black hole of mass M• dominates the gravitational potential inside an angular
‘radius of influence’ given by
θ• ∼ 0. 1
M•
106M
100km s−1
σ
2
1Mpc
D
(1.1)
Where θ• is the angular size projected on the sky of the radius of influence of
the black hole, σ is a typical velocity dispersion of stars in the galaxy and D is the
distance to the galaxy in Mpc. For typical nearby galaxies, this radius of influence
will be less than an arcsecond, so that the Hubble Space Telescope is required to make
the necessary observations.
The mass of the black hole may be sought by modeling the observed line of sight

30. 12
velocity distribution of the stars in a galaxy by superimposing collections of stellar
orbits and a range of different black hole masses to determine the best fitting model
velocity profile, using χ2
minimization techniques. The most reliable models must
include the Energy, Angular Momentum and Third-Integral of motion, and are there-
fore known as ‘three-integral’ models. While mass estimates obtained through these
means are often taken to be the most reliable, Valluri et al. (2004) remind us that we
do not know if orbits that physically exist in a galaxy are being selected to generate
the models, and that the errors in the fitting procedure are not normally distributed,
but more represent a range of equally acceptable values.
Finally, in what seems the most straightforward method at the outset, the mass
of a central black hole can be measured through the induced cusp in the rotation
velocities of gas in the nucleus of the galaxy (see Figure 1.7). State of the art models
do not account for dissipative effects or non-circular, coplanar motions and these
deviations from the circular thin disk model have important consequences for the
determination of black hole masses. Understanding these motions in the nuclei of
galaxies will allow further progress to be made in this field.
With the advent of HST, and the arrival of the necessary angular resolution to
probe scales of interest to black hole hunters, early emission line spectra of a few

31. 13
galaxies were obtained with the Faint Object Spectrograph (FOS). FOS was a single
aperture spectrograph and therefore rather inefficient for mapping velocity fields,
however early results were encouraging and showed that the kinematic signatures of
black holes could indeed be detected in the nuclei of some galaxies (Harms et al.,
1994; Ferrarese et al., 1996; van der Marel & van den Bosch, 1998; Ferrarese & Ford,
1999). The Space Telescope Imaging Spectrograph (STIS, see Brown et al., 2002) was
installed on board HST in 1997 making it possible to obtain long slit spectra of the gas
disks in galactic nuclei, and map out the velocity fields with much greater efficiency,
spatial coverage and resolution. Though the velocity fields may now be mapped much
more precisely, we will see later in this thesis that the observed velocity fields are not
those of uniform, circular disks, but rather pose more of a challenge to interpret.
We show the most reliable black hole determinations to date, as assessed by
Tremaine et al. (2002), in Table 1.2. The means of determination of each M• (by
those methods described briefly above) are also indicated. These data are plotted
in Figure 1.8 against the stellar velocity dispersion in the inner parts of each galaxy
(σc) showing the conspicuous correlation between these two parameters that was first
noted by Ferrarese & Merritt (2000) and Gebhardt et al. (2000a).
The scales on which σc are measured are much larger than the nuclear regions

32. 14
where the black hole dominates the potential, so the relationship implies a connection
between the formation and evolution of galaxies and the black holes they harbor.
Models describing this linked co-evolution, growing the systems that we observe today
from ‘seed’ black holes and small bulges, are beginning to provide some theoretical
frameworks to explain the observed correlations (Silk & Rees, 1998; Ostriker, 2000;
Haehnelt & Kauffmann, 2000; Adams et al., 2001). This type of co-evolution scenario
may also indicate that other connections between the central engine and large scale
host galaxy may exist.
1.4 This dissertation
In this dissertation we present investigations into a complete sample of nearby radio
galaxies, based primarily on spectroscopic observations obtained from the Hubble
Space Telescope. We set out to work towards answering the following questions:
• What is the nature of the spectra observed from the emission line gas in the
nuclei of nearby radio galaxies?
• What can be said about the masses of the black holes in these radio galaxies
based on the gas kinematics?
• How is the gas organized in the central regions?

33. 15
• What connections can be found between the central engines and that gas?
Understanding these matters in nearby galaxies, where we are able to probe on
scales that may be meaningful in terms of the physical processes that take place, may
allow us in future to gain insights into how AGN relate to each other, and how we
might make more meaningful interpretations of data we receive from more distant
sources where we do not have the resolution to probe the nuclei in such detail.
In Chapter 2, we will describe the selection of our sample of 21, nearby, radio-loud
elliptical galaxies, and summarize some previous observational work on the sample. In
Chapter 3, we describe the Hubble Space Telescope (HST) spectroscopic observations
we obtained of the sample galaxies, and present the resulting data set. In Chapter 4
we discuss work we have carried out on modeling the gas in the gravitational potential
of the galaxies, with and without black holes, and in Chapter 5 we go on to discuss
the global kinematic state of the nuclear gas, some properties of the central engines
and possible connections between the two. We present a summary of our conclusions
and a description of future directions for this work in Chapter 6.

34. 16
Table 1.1. Properties of various classes of Active Galactic Nuclei.
Type Broad Lines Narrow Lines Radio
(1) (2) (3) (4)
Radio-loud Quasars Loud
Radio-quiet Quasars Weak
Broad-line Radio Galaxies Loud
Narrow-line Radio Galaxies × Loud
BL-Lacs × × Loud
Seyfert-1 Weak
Seyfert-2 × Weak
LINERS × No
Note. — Col. (1): Name of the type of AGN; Cols. (2-3): Are broad and
narrow gas emission lines present?; Col. (4): Is the AGN radio loud?
References. — Adapted from Table 1.2 of Krolik (1999)

35. 17
Table 1.2. Reliable black hole mass measurements.
Name Type M• Method σ∗ Reference
(M ) (km s−1)
(1) (2) (3) (4) (5) (6)
Milky Way SBbc 1.8 × 106 s, p 103 (1)
M32 E2 2.5 × 106 s, 3I 75 (2)
M31 Sb 4.5 × 107 s 160 (3),(4),(5)
NGC 821 E4 3.7 × 107 s, 3I 209 (6),(7)
NGC 1023 SB0 4.4 × 107 s, 3I 205 (8)
NGC 1068 Sb 1.5 × 107 m 151 (9)
NGC 2778 E2 1.4 × 107 s, 3I 175 (6),(7)
NGC 2787 SB0 4.1 × 107 g 140 (10)
NGC 3115 S0 1.0 × 109 s 230 (11)
NGC 3245 S0 2.1 × 108 g 205 (12)
NGC 3377 E5 1.0 × 108 s, 3I 145 (6),(13)
NGC 3379 E1 1.0 × 108 s, 3I 206 (14)
NGC 3384 S0 1.6 × 107 s, 3I 143 (6),(7)
NGC 3608 E2 1.9 × 108 s, 3I 182 (6),(7)
NGC 4258 Sbc 3.9 × 107 m, a 130 (15)
NGC 4261 E2 5.2 × 108 g 315 (16)
NGC 4291 E2 3.1 × 108 s, 3I 242 (6),(7)
NGC 4342 S0 3.0 × 108 s, 3I 225 (17)
NGC 4459 S0 7.0 × 107 g 186 (10)
NGC 4473 E5 1.1 × 108 s, 3I 190 (6),(7)
NGC 4486 E0 3.0 × 109 g 375 (18),(19)
NGC 4564 E3 5.6 × 107 s, 3I 162 (6),(7)
NGC 4596 SB0 7.8 × 107 g 152 (10)
NGC 4649 E1 5.6 × 107 s, 3I 385 (6),(7)
NGC 4697 E4 1.7 × 108 s, 3I 177 (6),(7)
NGC 4742 E4 1.4 × 107 s, 3I 90 (20)
NGC 5845 E3 2.4 × 108 s, 3I 234 (6)
NGC 6251 E2 5.3 × 108 g 290 (21)
NGC 7052 E4 3.3 × 108 g 266 (22)
NGC 7457 S0 3.5 × 106 s, 3I 67 (6),(7)
IC 1459 E3 2.5 × 109 s, 3I 340 (23)
Note. — Data from Tremaine et al. (2002). Col. (1): Galaxy Name; Col. (2): Morphological Type; Col.
(3): Determined black hole mass; Col. (4): Method used, g - gas dynamics, m - maser dynamics, s - stars,
3I - three integral modeling, a - maser accelerations, p - proper motions; Col. (5): Central stellar velocity
dispersion; Col. (6): Reference for the Black Hole Mass
References. — Black Hole Mass References: (1) Chakrabarty & Saha (2001); (2) Verolme et al. (2002);
(3) Tremaine (1995); (4) Kormendy & Bender (1999); (5) Bacon et al. (2001); (6) Gebhardt et al. (2003); (7)
Pinkney et al. (2003); (8) Bower et al. (2001); (9) Greenhill & Gwinn (1997); (10) Sarzi et al. (2001); (11)
Kormendy et al. (1996); (12) Barth et al. (2001); (13) Kormendy et al. (1998); (14) Gebhardt et al. (2000b);
(15) Herrnstein et al. (1999); (16) Ferrarese et al. (1996); (17) Cretton & van den Bosch (1999); (18) Harms
et al. (1994); (19) Macchetto et al. (1997); (20) Kaiser, in preparation; (21) Ferrarese & Ford (1999); (22)
van der Marel & van den Bosch (1998); (23) Cappellari et al. (2002).

36. 18
Core
Optical galaxy
Jet
Jet
Lobe
Lobe
Figure 1.1 A cartoon view of a radio galaxy showing the core, jets and lobes as ob-
served at radio wavelengths, a typical size scale for the elliptical host galaxy observed
at visible wavelengths is also indicated.

37. 19
Figure 1.2 An example of a FR-I (left) and a FR-II (right) radio galaxy, observed at
20 cm wavelengths by Owen & Ledlow (1997). The FR-I is characterized by bright
jets, closest to the nucleus, while the FR-II has bright radio lobes.

38. 20
Jet
Jet
Clouds
Broad Line
Accretion Disk
Black Hole
Obscuring TorusObscuring Torus
Narrow Line Region
Figure 1.3 A representation of the unified scheme for the central engine of active
galactic nuclei (AGN); different classes of AGN would be observed by viewing the
model from different angles. (after Urry & Padovani, 1995).

39. 21
Figure 1.4 From Martel et al. (2000): The dust disk in the nucleus of NGC 383
showing intriguing morphological structures in the dust distribution.

40. 22
Figure 1.5 From Wada (2004), the segregation of various types of AGNs from the
point of view of the circumnuclear gas disk.

41. 23
Figure 1.6 The famous cartoon by Phinney (1994) capturing the essence of the fueling
problem in AGN: getting the food with the large (angular momentum) spoon into
the small (area and angular momentum) mouth.

42. 24
Figure 1.7 Cusp in velocities in the nuclear region of M87. Dotted line indicates the
continuum flux. (figure from Macchetto et al., 1997).

43. 25
60 70 80 90100 200 300 400
σc (km s-1
)
106
107
108
109
1010
MBH(MSun)
Figure 1.8 Measured black hole masses (MBH ) as a function of central stellar veloc-
ity dispersion (σc) for reliable determinations summarized by Tremaine et al. (2002).
The line indicates the fit M• = 1.3 × 108
M (σc/200)4.72
. (*) indicates masses de-
termined from stellar dynamical modeling; (◦) indicates masses determined from gas
kinematics; and ( ) indicates masses determined from MASER kinematics.

44. Chapter 2
The UGC FR-I Sample
In this Chapter we introduce the UGC FR-I galaxy sample, which is the central
sample of galaxies discussed in this dissertation. We briefly discuss the radio and
optical properties of the sample, and present some stellar kinematical data which
relates to relationships between black hole mass and properties of the host galaxies.
2.1 Sample selection and properties
Our galaxy sample (the UGC FR-I sample) contains all 21 nearby (vr < 7000 km s−1
),
elliptical or S0 galaxies in the declination range −5◦
< δ < 70◦
in the UGC catalog
(Nilson, 1973, limits magnitude mB < 14.m
6 and angular size θp > 1.0) that are
extended radio-loud sources (larger than 10 at 3σ on VLA A-Array maps, which
crudely ensures that the sources are extended, and brighter than 150 mJy from single
26

45. 27
dish flux density measurements at 1400 MHz). The source information is shown in
Table 2.1. The selection criteria result in a complete sample of nearby radio galaxies
with jets.
This complete sample was drawn from a catalog of 176 radio-loud galaxies con-
structed by Condon & Broderick (1988), by position coincidence of radio identifica-
tions in the Green Bank 1400 MHz sky maps and galaxies in the UGC catalog. All
of the galaxies fall into the Fanaroff & Riley (1974) Type-I (FR-I) radio classification
(see Xu et al., 2000, for a description of the radio properties of our sample); i.e. they
are low luminosity radio galaxies, with jets that are brightest nearest to the nucleus.
(In contrast, FR-II galaxies are more powerful radio sources, and have bright spots
at the far edges of their radio lobes, the only FR-II within 7000 km s−1
is the radio
galaxy Cen-A, which is in the southern sky.)
The combination of selecting extended sources and early type galaxies results in
the primary energy source of each galaxy in the sample falling into ‘monster’ rather
than ‘starburst’ classification of Condon & Broderick (1988) based on their infra-red
to radio flux ratios (for example, see Heckman et al., 1983)
u ≡ log
S60µm
S1400MHz
1.6, (2.1)

46. 28
and infra-red spectral gradients
αIR ≡
log (S60µm/S25µm)
log (60/25)
< +1.25. (2.2)
2.2 Multiwavelength observations
Prior to the work presented in this dissertation, imaging of the galaxies had been car-
ried out at various wavelengths with a variety of instruments. The fluxes of the UGC
FR-I nuclei in various wave-bands are given in Table 2.2. Particular attention has
been paid to the sample at radio and optical wavelengths, and we briefly summarize
those results below. We describe spectroscopic observations of the nuclei, that we
obtained using HST, in Chapter 3. In Figures 2.1 to 2.21 we show continuum optical
images of the central region of each galaxy – indicating the directions of the galaxy
major axes and radio jet axes, along with the positions of the spectroscopic slits (see
Chapter 3). The orientations of the slits compared to morphological features of the
nuclei are explained in the individual description of the observations given in §3.4 and
offsets are given in Table 3.3. For the majority of cases the STIS slits were aligned
within 10◦
of the galaxy major axis.

47. 29
2.2.1 Radio properties
Radio observations of the sample galaxies were obtained using both the Very Large Ar-
ray (VLA; Wrobel, et al., in preparation) and the Very Large Baseline Array (VLBA),
descriptions of the radio morphology are given for each galaxy in §3.4. Xu et al. (2000)
report the results of a program using the VLBA to observe 17 of the UGC FR-I sam-
ple galaxies at 1.67 GHz. At a resolution of ∼ 10 × 4 mas, five galaxies showed only
an unresolved radio core, 10 galaxies showed core-jet structures, and two galaxies
showed twin-jet structures. Comparing the VLBA jets (on parsec scales) to the VLA
jets (on kiloparsec scales) they found that the VLBA and VLA jets are well aligned
and that the jet-to-counterjet surface brightness ratios, or the sidedness, decreases
systematically with increasing distance along the jet. Xu et al. (2000) attribute the
sidedness of the jets to the Doppler boosting effect, and its decline to the deceleration
of the jets.
2.2.2 Optical properties
A photometric analysis of the nuclei of the UGC FR-I sample galaxies was performed
by Verdoes Kleijn et al. (1999), based on observations made using the WFPC2 in-
strument on board HST (the photometric analyses of UGC 7115 and UGC 12064 are

48. 30
presented in Appendix A of Verdoes Kleijn et al., 2002a). Verdoes Kleijn et al. (1999)
obtained V - and I-band images and narrow-band images centered on the Hα + [N II]
emission lines. They found that although obscuration by dust prevents satisfactory
determinations of the central cusp slopes, that the data suggest that most of the
sample galaxies have shallow cores.
Dust is detected in all but two galaxies and central emission line gas is detected in
all of the galaxies in the sample. There are a wide variety of central dust morphologies,
ranging from central disks to lanes and irregular distributions; the dust morphologies
for each individual source are described in §3.4.
2.3 Stellar Dynamics
We estimate the central stellar velocity dispersion (σc) within one eighth of the effec-
tive radius (re/8) using the relationship
σap
σc
=
rap
re/8
−0.04
, (2.3)
as described by Jørgensen et al. (1995) and basing σap on ground based measure-
ments from various sources as referenced in Table 2.3.
We found the effective radii of the sample nuclei using the WFPC2 imaging de-

49. 31
scribed by Verdoes Kleijn et al. (1999), and fitting a r1/4
-law profile of the form
I = I0 exp (−7.67(x/re)0.25
− 1) (2.4)
to the stellar surface brightness profile outside of the break radius.
This procedure to find σc follows the same prescription as Ferrarese & Merritt
(2000). Using their relationship between σc and M• (Merritt & Ferrarese, 2001):
M• = 1.30 × 108
M
σc
200kms−1
4.72
, (2.5)
we estimate values of M• for our sample galaxies. The stellar velocity dispersions
and black hole mass estimates are listed in Table 2.3, where we also give measured
black hole masses for the 5 sample galaxies for which such data exist. The five
measured black hole masses are plotted on a M• − σc diagram including all reliable
black hole masses (cf Figure 1.8) in Figure 2.22.
Variations of this relationship give different values for M•, however for our pur-
poses in this dissertation we will give black hole mass estimates based on the above
relationship and remind the reader that the masses may vary somewhat from these
values.

50. 32
Table 2.1. Properties of the galaxy sample members.
NGC UGC Other Names Type vsys STIS Scale MB log(L1400) Axis Ratio
(km s−1) (pc/pixel) (mag) (W Hz−1) b/a
(1) (2) (3) (4) (5) (6) (7) (8) (9)
193 408 E-S0 4342.5 15.0 -21.0 23.93 (0.18)
315 597 E 5092.5 17.6 -22.6 24.10 0.23
383 689 3C 31 E-S0 4890.0 16.9 -22.2 24.51 0.77
541 1004 E 5497.5 19.0 -21.7 23.94 0.91
741 1413 E 5265.0 18.2 -22.6 23.85 · · ·
1841 3C 66B E 6360.0 22.0 -22.5 24.94 ∼ 0.98
2329 3695 E-S0 5725.0 19.8 -21.9 23.76 0.68
2892 5073 E 6810.0 23.6 -21.1 23.36 · · ·
3801 6635 S0 3255.0 11.3 -20.8 23.49 (0.12)
3862 6723 3C 264 E 6330.0 21.9 -21.7 24.75 ∼ 0.99
7115 E 6787.5 23.5 -21.0 23.94 ∼ 0.95
4261 7360 3C 270 E 2212.5 7.7 -21.5 24.40 0.46
4335 7455 E-S0 4672.5 16.2 -21.6 23.11 0.41
4374 7494 M84, 3C 272.1 S0 1155.0 4.0 -20.9 23.35 (0.15)
4486 7654 M87, 3C 274 E 1155.0 4.0 -22.2 24.90 · · ·
5127 8419 E 4830.0 16.7 -21.3 24.08 (0.25)
5141 8433 S0 5302.5 18.4 -21.0 23.80 (0.25)
5490 9058 E 5790.0 20.1 -21.7 23.79 (0.35)
7052 11718 E 4155.0 14.4 -21.0 23.04 0.30
12064 3C 449 E-S0 5122.5 17.7 -20.8 24.38 0.54
7626 12531 E 3495.0 12.1 -21.7 23.37 (0.17)
Note. — Col. (1): NGC number where available; Col. (2): Upsalla General
Catalog (UGC) number; Col. (3): Alternative names; Col. (4): From the NASA
Extragalactic Database (NED); Col. (5): Measured from the stellar kinematics
(from NED); Col. (6): Parsecs per unbinned STIS pixel; Col. (7): Absolute blue
magnitude from Condon & Broderick (1988); Col. (8): Radio Luminosity from
Condon & Broderick (1988); Col. (9): Dust disk axis ratio from Verdoes Kleijn
et al. (1999), numbers in parentheses indicate a dust lane where the width:length
ratio is given instead.

51. 33
Table 2.2. Multiwavelength Fluxes of the UGC FR-I Sample Galaxies.
Galaxy Vnuc Inuc VLBApeak VLApeak X-raysoft X-rayhard
(mJy/b.a.) (mJy/b.a.) (erg cm−2 s−1) (erg cm−2 s−1)
(1) (2) (3) (4) (5) (6) (7)
NGC 193 6.20 × 10−18 1.10 × 10−17 29.8 40.0 · · · · · ·
NGC 315 3.30 × 10−17 4.70 × 10−17 224 396 4.63 × 10−13 9.61 × 10−13
NGC 383 2.40 × 10−17 1.80 × 10−17 44.1 89.0 3.91 × 10−14 6.78 × 10−14
NGC 541 7.50 × 10−18 7.40 × 10−18 1.90 8.00 · · · · · ·
UGC 1841 5.20 × 10−17 3.70 × 10−17 112 131 2.37 × 10−13 8.79 × 10−14
NGC 2329 1.70 × 10−16 1.20 × 10−16 49.7 117 · · · · · ·
NGC 2892 1.60 × 10−17 1.40 × 10−17 15.3 22.0 · · · · · ·
NGC 3862 1.90 × 10−16 1.40 × 10−16 123 386 · · · · · ·
UGC 7115 3.40 × 10−17 3.40 × 10−17 · · · · · · · · · · · ·
NGC 4261 3.80 × 10−18 1.00 × 10−17 67.4 165 9.64 × 10−14 9.86 × 10−13
NGC 4335 2.50 × 10−17 3.00 × 10−17 9.20 15.0 · · · · · ·
NGC 4374 6.20 × 10−17 7.30 × 10−17 106 112 1.89 × 10−13 8.94 × 10−14
NGC 4486 6.40 × 10−16 3.20 × 10−16 1570 3600 · · · · · ·
NGC 5127 · · · · · · 3.90 7.00 · · · · · ·
NGC 5141 9.50 × 10−18 1.40 × 10−17 35.5 71.0 · · · · · ·
NGC 5490 6.70 × 10−19 4.00 × 10−18 20.0 41.0 · · · · · ·
NGC 7052 · · · · · · 2.79 × 101 36.0 · · · · · ·
UGC 12064 · · · 3.10 × 10−17 · · · · · · · · · · · ·
NGC 7626 3.70 × 10−18 1.10 × 10−17 12.5 23.0 · · · · · ·
Note. — Col. (1): Galaxy Name; Cols. (2-3): The nuclear point source V and
I fluxes (from WFPC2 imaging) in (erg s−1
cm−2 ˚A−1
); Cols. (4-5) Peak fluxes in
the radio core from VLBA and VLA observations; Cols. (5-6) Unabsorbed X-ray
fluxes from the nucleus in soft (0.02 - 2 keV) and hard (2 - 10 keV) bands from
Chandra observations.
References. — Cols. 2-3: Verdoes Kleijn et al. (2002a); Cols. 4-5: Xu et al.
(2000); Cols. 6-7 E. Colbert (Private Communication).

52. 34
Table 2.3. Stellar velocity dispersions and estimated black hole masses.
Galaxy σmeas Ref. rap re σc M• MF01 M•Meas. Ref.
(km s−1) (arcsec) (arcsec) (km s−1) (108M ) (108M )
NGC 193 · · · · · · · · · 25.90 · · · · · · · · · · · ·
NGC 315 357 (1) 1.48 62.78 334 14.6 · · · · · ·
NGC 383 311 (1) 1.48 22.65 303 9.24 · · · · · ·
NGC 541 225 (1) 1.48 21.34 220 2.03 · · · · · ·
NGC 741 284 (1) 1.48 47.01 269 5.24 · · · · · ·
UGC 1841 382 (10) Re · · · 352 18.6 · · · · · ·
NGC 2329 274 (1) 1.48 26.81 265 4.92 · · · · · ·
NGC 2892 · · · · · · · · · 33.51 · · · · · · · · · · · ·
NGC 3801 225 (2) Re? · · · 207 1.53 · · · · · ·
NGC 3862 263 (1) 1.48 12.87 264 4.66 · · · · · ·
UGC 7115 198 (3) 3.39 13.63 204 1.41 · · · · · ·
NGC 4261 306 (1) 1.48 38.25 292 7.75 5.4 (5)
NGC 4335 282 (9) 2Re · · · 252 3.9 ≤ 1.0 (9)
NGC 4374 304 (1) 1.48 43.87 288 7.30 4.3 (6)
NGC 4486 383 (1) 4.50 113.87 366 22.5 35.7 (7)
NGC 5127 198 (3) 3.39 48.12 194 1.11 · · · · · ·
NGC 5141 · · · · · · · · · 15.13 · · · · · · · · · · · ·
NGC 5490 305 (1) 1.48 15.39 302 9.07 · · · · · ·
NGC 7052 270 (4) ∼ Re · · · 248 3.62 3.7 (8)
UGC 12064 233 (10) Re · · · 214 1.80 · · · · · ·
NGC 7626 244 (1) 1.48 34.16 234 2.72 · · · · · ·
Note. — Measured stellar velocity dispersions are shown, with the source
reference indicated. These have then been adapted to yield σc based on mea-
surements of re from our WFPC2 data (see text). The relation of Merritt &
Ferrarese (2001) was then used to yield estimates of the black hole masses.
References. — (1) Davies et al. (1987); (2) Di Nella et al. (1995); (3) Tonry
& Davis (1981); (4) Wagner et al. (1988); (5) Ferrarese et al. (1996); (6) Ma-
ciejewski & Binney (2001); (7) Macchetto et al. (1997); (8) van der Marel & van
den Bosch (1998); (9) Verdoes Kleijn et al. (2002b); (10) Balcells et al. (1995);
(11) van der Marel & van den Bosch (1998).

53. 35
0.5"
Figure 2.1 Optical continuum image of the nuclear region of NGC 193. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

54. 36
0.5"
Figure 2.2 Optical continuum image of the nuclear region of NGC 315. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

55. 37
0.5"
Figure 2.3 Optical continuum image of the nuclear region of NGC 383. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

56. 38
0.5"
Figure 2.4 Optical continuum image of the nuclear region of NGC 541. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

57. 39
0.5"
Figure 2.5 Optical continuum image of the nuclear region of NGC 741. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

58. 40
0.5"
Figure 2.6 Optical continuum image of the nuclear region of UGC 1841. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

59. 41
0.5"
Figure 2.7 Optical continuum image of the nuclear region of NGC 2329. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

60. 42
0.5"
Figure 2.8 Optical continuum image of the nuclear region of NGC 2892. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

61. 43
0.5"
Figure 2.9 Optical continuum image of the nuclear region of NGC 3801. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

62. 44
0.5"
Figure 2.10 Optical continuum image of the nuclear region of NGC 3862. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

63. 45
0.5"
Figure 2.11 Optical continuum image of the nuclear region of UGC 7115. The position
of the spectroscopic slit used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

64. 46
0.5"
Figure 2.12 Optical continuum image of the nuclear region of NGC 4261. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

65. 47
0.5"
Figure 2.13 Optical continuum image of the nuclear region of NGC 4335. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

66. 48
0.5"
Figure 2.14 Optical continuum image of the nuclear region of M84. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

67. 49
0.5"
Figure 2.15 Optical continuum image of the nuclear region of NGC 4486. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

68. 50
0.5"
Figure 2.16 Optical continuum image of the nuclear region of NGC 5127. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

69. 51
0.5"
Figure 2.17 Optical continuum image of the nuclear region of NGC 5141. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

70. 52
0.5"
Figure 2.18 Optical continuum image of the nuclear region of NGC 5490. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

71. 53
0.5"
Figure 2.19 Optical continuum image of the nuclear region of NGC 7052. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

72. 54
0.5"
Figure 2.20 Optical continuum image of the nuclear region of UGC 12064. The
position of the spectroscopic slits used in the observing program described in Chapter
3 are indicated. The dashed line indicates the galaxy major axis position angle, the
dotted line indicates the direction of the radio jet. A north-east indicator is shown.

73. 55
0.5"
Figure 2.21 Optical continuum image of the nuclear region of NGC 7626. The position
of the spectroscopic slits used in the observing program described in Chapter 3 are
indicated. The dashed line indicates the galaxy major axis position angle, the dotted
line indicates the direction of the radio jet. A north-east indicator is shown.

74. 56
60 70 80 90100 200 300 400
σc (km s-1
)
106
107
108
109
1010
MBH(MSun)
Figure 2.22 Measured black hole masses (MBH ) as a function of central stellar velocity
dispersion (σc) for reliable determinations summarized by Tremaine et al. (2002). The
line indicates the fit M• = 1.3 × 108
M (σc/200)4.72
. (*) indicates masses determined
from stellar dynamical modeling; (◦) indicates masses determined from gas kinemat-
ics; and ( ) indicates masses determined from MASER kinematics. The members of
the UGC FR-I sample which have previous black hole mass measurements are shown
as black squares ( ).

75. Chapter 3
STIS spectroscopy of the emission
line gas in the nuclei of nearby
FR-I galaxies
This chapter originally appeared as part of Noel-Storr et al. (2003).
3.1 Introduction
We present the results of the analysis of a set of medium resolution spectra, obtained
by the Space Telescope Imaging Spectrograph on board the Hubble Space Telescope,
of the emission line gas present in the nuclei of a complete sample of 21 nearby, early-
type galaxies with radio jets (the UGC FR-I Sample). For each galaxy nucleus we
present spectroscopic data in the region of Hα and the derived kinematics.
57

76. 58
We find that in 67% of the nuclei the gas appears to be rotating and, with one
exception, the cases where rotation is not seen are either face on or have complex
central morphologies. We find that in 62% of the nuclei the fit to the central spectrum
is improved by the inclusion of a broad component. The broad components have a
mean velocity dispersion of 1349 ± 345 km s−1
and are redshifted from the narrow
line components (assuming an origin in Hα) by 486 ± 443 km s−1
.
The chapter is organized as follows. In section 2 we describe the sample and in
section 3 we describe the STIS spectroscopic observations and the data reduction. In
section 4 we describe our analysis procedures. We present our initial interpretations
in section 5 and draw conclusions in section 6. We use a Hubble constant of H0 =
50km s−1
Mpc−1
throughout.
3.2 STIS Observations
In program 8236 we used HST/STIS (see Kimble, et al., 1998) to take spectra of the
19 sample members not previously or concurrently observed by others. Observations
were carried out at both medium (for 19/19 galaxies) and low (for 4/19 galaxies) res-
olution, with the G750M and the G430L and G750L gratings respectively. Our medium
resolution observation log is shown in Table 3.1. We will present the low resolution

77. 59
spectra in a future paper. We include in our analysis similar medium resolution data
obtained by R. Green and collaborators for the nucleus of M84 (program 7124, see
Bower et al., 1998) and H. Ford and collaborators for the nucleus of M87 (program
8666) to complete the data set for UGC FR-I galaxies.
In our program, we observed each galaxy in three parallel, adjacent slit positions.
For each slit position we obtained two exposures with a shift of 0.202800 (4 unbinned
STIS Pixels) along the slit direction to enable us to more efficiently remove detector
effects (bad pixels, etc.). In the case of UGC 7115 we observed in only one slit position
- we sacrificed STIS observing time to make WFPC2 observations of this galaxy as
it had not been included in our earlier WFPC2 program.
The data for M87 (NGC 4486) were obtained in a similar manner to our own.
In the case of M84 (NGC 4374) the observation pairs were not shifted along the slit
direction, thus some detector effects may remain, though will be much less significant
thanks to the far greater signal to noise.
We list the instrumental properties of the STIS configurations used in Table 3.2.
Panel (a) of Figures 3.2 to 3.22 (see key in Figure 3.1) shows the location of the STIS
slits on each galaxy observed, along with the position angles of the galaxy major axes
and radio-jet axes. For the majority of cases the STIS slits were aligned within 10◦

78. 60
of the galaxy major axes (Table 3.3), with the exceptions that we note below.
In the cases of NGC 741 [∆PA(gal, slit) = 19.4◦
] and NGC 2892 [∆PA(gal, slit) =
22◦
] more freedom in orientation was allowed to provide reasonable observing win-
dows.
In NGC 3862 and UGC 7115 the position angles of the galaxy major axes are
hard to determine, we decided to position the slits approximately perpendicular to
the radio jets [∆PA(jet, slit) = 92.3◦
& 110.4◦
, respectively]. The central major axis
is also hard to determine in NGC 541, where the central isophotes rotate consid-
erably. In this case we chose to use a slit position somewhere around the mean of
the isophotal position angles with considerable leeway given to allow for reasonable
observing windows.
For UGC 12064 the slits were aligned along the major axis of the prominent dust
disk, which is offset from the galaxy major axis by ∼ 50◦
.
The slits in M84 were positioned approximately perpendicular to the radio jet,
which lies close to the major axis of the nuclear gas. For NGC 4486 (M87) the slits
were positioned to follow certain morphological structures across the nuclear regions.

79. 61
3.2.1 Data Reduction
We used the standard STIS calibration pipeline (calstis, see Brown et al., 2002) to
perform bias, dark and flat-field corrections using the best available reference files.
We used calstis version 2.13 (26-April-2002) throughout the data reduction1
.
We shifted the rows of alternating observations by 4 pixels, so that they were
properly aligned with their counterparts and combined them using the STSDAS rou-
tine ocrreject. We cleaned co-incident cosmic rays and negative bad pixels, which
would not be caught by ocrreject, using the NOAO/IRAF task cosmicrays. At
each step we carefully investigated the effects of varying task parameters to insure we
were not damaging valid data while removing most cosmic rays.
We made use of the STIS calibration pipeline tasks wavecal and x2d to perform
wavelength calibration and image rectification respectively. The error introduced by
rectifying after shifting one of the images is ∼< 0.05 pixels (which is ∼< 1.3 km s−1
at
6750˚A).
Panel (b) of Figures 3.2 to 3.22 (see key in Figure 3.1) shows (i) the central strip
1
We found that significant variations in measured parameters could be introduced by using differ-
ent calstis versions. Versions 2.4 and 2.13 produce consistent results, while intermediate versions
do not.

80. 62
of the reduced spectrum for each slit position observed on each galaxy, along with
(ii) Gaussian line fits to the same (see §4.1).
3.3 Analysis
In this section we describe the line fitting that we carried out on each spectral row
of each reduced CCD spectral image, firstly with a single Gaussian per spectral line
(§4.1) and secondly with the inclusion of an additional free component (§4.2). In
§4.3 we discuss the sizes of the errors on quoted parameters from various sources. In
§4.4 we describe each of the UGC FR-I sample members in turn. For each galaxy
we define the central spectrum as the row with the greatest integrated flux after the
data reduction. We list the row numbers in the final x2d image corresponding to the
central spectrum in Table 3.1.
3.3.1 Single Gaussian line fitting
In the G750M spectra we expect to find the emission lines in the vicinity of Hα that are
listed in Table 3.4. We used wavelengths from the recent measurements of Wallerstein
et al. (2001) and converted from air to vacuum wavelengths using the IAU standard
formula

81. 63
λvac − λair
λair
= 6.4328 × 10−5
+
2.94981 × 10−2
146 − (104/λair)2 +
2.5540 × 10−4
41 − (104/λair)2 . (3.1)
Using one Gaussian to represent each of these five lines we obtain a set of 7 free
parameters to fit: the continuum flux level, velocity (vr), velocity dispersion (σ), and
the fluxes of each line. The flux of [N II]6550 was fixed in a ratio of 1:3 with the
flux of [N II]6585 based on the transition probabilities derived from atomic physics
(Osterbrock, 1989).
We used a χ2
minimization routine (using Levenberg-Marquardt iterations, see
Press et al., 1992) to fit the Gaussian template to the observed spectra. The applica-
tion of this fitting technique and development of this routine are described by van der
Marel & van den Bosch (1998). Formal errors are drawn from the covariance matrix
of the fit. As we do not expect the noise in each spectrum to be normally-distributed
after the steps of wavelength calibration and two-dimensional rectification, these error
values should be treated strictly as the formal fit errors under the understanding that
the size of the real errors may be somewhat different (see §4.3 below).
From this point on we only consider data points where the formal errors from
the fits meet the following criteria, allowing us to exclude unreliable data points
originating from poorly constrained fits:

82. 64
∆σ < 50 km s−1
(3.2)
∆F(Hα)
F(Hα)
< 0.75 (3.3)
where ∆σ and ∆F(Hα) are the errors in velocity dispersion and line flux respec-
tively. The main constraint arises from the limit on the velocity dispersion error. The
very large flux error allowed is in place to remove only the few remaining bad data
points where the profile very precisely fits the noise.
In Panel (e) of Figures 3.2 to 3.22 (see key in Figure 3.1) we show profiles of (i)
radial velocity, (ii) velocity dispersion, (iii) [N II]6585 line flux and (iv) [N II] / Hα
ratios resulting from this fitting procedure for each of our sample galaxies. These
profiles are combined and visualized in 2D for each galaxy in Panel (c) of Figures 3.2
to 3.22.
We present the fit data in Tables 3.5 to 3.25, where the errors given are the formal
errors from the fit. In these tables Column (1) is the row number of the portion of the
spectrum fitted. Column (2) shows the offset along the slit direction in arcseconds
from the row with the greatest integrated flux. Columns (3) and (4) give the radial
velocities (vr) and gas velocity dispersions (σgas) respectively. Column (5) gives the

83. 65
line flux of the Hα line and columns (6) and (7) show its ratio against the fluxes of
the [N II]6585 and [S II]total (the total flux of the two [S II] lines) respectively. Column
(8) gives the reduced χ2
(R2
) value of the resulting fit.
We repeated the fit for the central row of the galaxy NGC 4335 varying the set of
free parameters in order to estimate the reliability of the fits that we had used. We
found that the fit was stable to within one formal error on all quoted parameters when
the velocities, velocity dispersions and fluxes of all parameters were fit independently.
The signal to noise falls off rapidly outside of the very nuclear regions so it is not
possible to consistently run fits with a large number of free parameters. The results
for the nucleus of NGC 4335 satisfy us that we are justified in fixing the parameters
in the manner that we have chosen, without adding any obvious biases to our results.
3.3.2 Fits with an additional free component
In many cases, as the very central pixels are reached the fit begins to do a poorer
job of matching the observed profile. In an attempt to improve the fit to the narrow
centers of the lines we tested a fit for the central spectrum of each galaxy including
an additional fit component with independent velocity, velocity dispersion and flux,
along with the original set of five Gaussians. The fits to the central spectra are
shown in Panel (d) of Figures 3.2 to 3.22 (see key in Figure 3.1), (i) excluding and

84. 66
(ii) including the additional component.
We assessed the effectiveness of including this component in each galaxy based
on (1) an improvement in the mean of the absolute value of the residuals from the
fit ≥ 5% (2) an improvement in the reduced χ2
value of the fit such that (R2
1 −
R2
2)/R2
1 ≥ 0.15 and (3) an improvement judged by eye in the fit compared to the
data. We assigned a score to each galaxy, with one point available for each of the
three categories. We consider scores of 2 or 3 to be indicative of the presence of a
broad component, a score of 1 indicates the possibility of a broad component, while
we treat a score of 0 as a none detection. The three parameters and scores are listed
in Table 3.27.
We find an additional free component improves the fit in 62% (N = 13) of the
sample galaxies. In the cases of NGC 2329 and NGC 3862 the component appears to
represent a non-flat continuum. The kinematic parameters for each galaxy including
the additional free component are listed in Table 3.28 and the flux parameters in
Table 3.29. We present further interpretation of the nature and origin of the features
fit by the additional free component in §5.3.

85. 67
3.3.3 Quantifying error sources
The STIS data handbook (Brown et al., 2002) gives the following absolute and relative
accuracies applicable to this work: A wavelength absolute calibration error (∆λ offset)
of 0.1 to 0.3 pixels (2.6 to 7.7 km s−1
at 6500˚A) within an exposure, and from 0.2 to
0.5 pixels (5.1 to 12.8 km s−1
at 6500˚A) between exposures. An absolute photometry
error of 5% and a relative photometry error of 2% within a single exposure assuming
a wide slit observation. 5 µm variations in slit width along the slit lengths could
result in variations of up to 20% in flux along the 0.1 slit.
In Verdoes Kleijn et al. (2002a) Hα + [N II] fluxes were presented for each nucleus
in the sample. The values presented there agree well with the values we find here,
certainly given our limited ability to extract comparable apertures and within the
20% potential flux errors noted above.
In Section 1 we indicated that a 1.3 km s−1
error could be incorporated into the
final data as a result of shifting the spectra for image combination and cosmic ray
rejection. This shift is insignificant compared to other error sources.
In Table 3.26 we showed that by allowing different free parameters within the
single-Gaussian-per-line fit produced changes in the measurement in velocity of ∼

86. 68
8 km s−1
and of ∼ 16 km s−1
in velocity dispersion for the nucleus of NGC 4335.
In Table 3.32 we show the effect on the measured velocities and velocity dispersions
of the various components for each of the models described in the previous section,
again for the case of NGC 4335. This illustrates that the measured velocities of the
narrow lines may vary by up to ∼ 20 km s−1
(and velocity dispersions by as much as
∼ 110 km s−1
) when additional components in the line shape are taken into account.
In the nuclei of NGC 383 and NGC 4335 (representing cases with blended and
less blended lines respectively) we repeated the narrow line fit to the central spec-
trum with 286 different combinations of input velocity and velocity dispersion; vary-
ing the velocities over a range of 2000 km s−1
and the velocity dispersions over a
range of 6000 km s−1
. In the case of NGC 4335 we found that the velocity varied
by ±10.91 km s−1
and the velocity dispersion by just ±0.02 km s−1
. In the case of
NGC 383 we found that the velocity varied by ±13.04 km s−1
and the velocity disper-
sion by just ±3.76 km s−1
. There was a systematic effect relating input and output
velocities in both cases.
We conclude that reasonable estimates of the genuine errors on each of our mea-
sured parameters are: 5% - 10% on fluxes (dominated by the effects of variations along
the narrow slits and the STIS absolute calibration); and ∼ 20 km s−1
on velocities

87. 69
and velocity dispersions (dominated by the fit model dependency of the results).
3.4 Individual galaxy descriptions
Below, we give descriptions of each member of the UGC FR-I sample in turn. The
galaxy classifications are taken from the NASA Extragalactic Database, which lists
references in which the terms used are described. Descriptions of dust properties and
radio sources are as presented by Verdoes Kleijn et al. (1999) and Xu et al. (2000)
respectively.
NGC 193 (UGC 408) This S0 galaxy has a complex gas morphology with two
lanes apparent in the central regions (the most clearly defined lane has a width :
length = 0.18). It has a core-jet radio morphology on VLA and VLBA scales. The
STIS slits were aligned parallel to the galaxy major axis. The central kinematic and
flux properties are listed in Table 3.5; the gas does not exhibit a regular rotation
curve, though it does appear dominated by systematic rather than random motions.
The fit to the central spectrum is improved by the addition of a broad component.
Data for this galaxy are shown in Figure 3.2 (see key in Figure 3.1 for an explanation
of these plots).

88. 70
NGC 315 (UGC 597) This elliptical galaxy has a nuclear dust disk (b/a = 0.23).
It has a core-jet radio morphology on VLA and VLBA scales. The STIS slits were
aligned parallel to the galaxy major axis. The central kinematic and flux properties
are listed in Table 3.6; the gas appears to be in organized motion, possibly regular
rotation. The fit to the central spectrum is improved by the addition of a broad
component. Data for this galaxy are shown in Figure 3.3 (see key in Figure 3.1 for
an explanation of these plots).
NGC 383 (UGC 689) This S0 galaxy has a nuclear dust disk (b/a = 0.77). It
has a core-jet radio morphology on VLBA scales, and a twin-jet morphology on VLA
scales. The STIS slits were aligned parallel to the galaxy major axis. The central
kinematic and flux properties are listed in Table 3.7; the gas exhibits a regular rotation
profile. In the negative offset side slit there is a dip in the velocity dispersion profile
at a position close to the nucleus. The fit to the central spectrum is improved by the
addition of a broad component. Data for this galaxy are shown in Figure 3.4 (see key
in Figure 3.1 for an explanation of these plots).
NGC 541 (UGC 1004) This cD S0 galaxy has a nuclear dust disk (b/a = 0.91). It
has a radio core on VLBA scales and a core-jet morphology on VLA scales. The STIS

89. 71
slits were aligned to a mean of the position angles of the central isophotes measured
from our WFPC/2 images, which vary considerably. We allowed considerable flexibil-
ity in position angle to enable reasonable observing windows. The central kinematic
and flux properties are listed in Table 3.8; the gas does not exhibit a regular rotation
profile. The fit to the central spectrum is not significantly improved by the addition
of a broad component, though the fit improves somewhat when judged by eye. Data
for this galaxy are shown in Figure 3.5 (see key in Figure 3.1 for an explanation of
these plots).
NGC 741 (UGC 1413) This E0 galaxy has no apparent nuclear dust. It has a
radio core on VLBA scales and a core-jet morphology on VLA scales. The STIS slits
were aligned approximately parallel to the galaxy major axis, however a certain degree
of freedom was allowed in slit placement to allow reasonable observing windows. The
central kinematic and flux properties are listed in Table 3.9. Very few points had
sufficient signal to noise to obtain good fits in these data, it has not been included
in further analysis of global kinematic properties. The fit to the central spectrum is
not improved by the addition of a broad component. Data for this galaxy are shown
in Figure 3.6 (see key in Figure 3.1 for an explanation of these plots).

90. 72
UGC 1841 This elliptical galaxy has a nuclear dust disk (b/a ∼ 0.98). It has a
core-jet radio morphology on VLBA and VLA scales. The STIS slits were aligned
parallel to the galaxy major axis. The central kinematic and flux properties are listed
in Table 3.10; the gas does not exhibit a regular rotation profile. The fit to the central
spectrum is improved by the addition of a broad component. Data for this galaxy
are shown in Figure 3.7 (see key in Figure 3.1 for an explanation of these plots).
NGC 2329 (UGC 3695) This S0 galaxy has a nuclear dust disk (b/a = 0.68).
It has a core-jet radio morphology on VLBA and VLA scales. The STIS slits were
aligned parallel to the galaxy major axis. The central kinematic and flux properties
are listed in Table 3.11; the gas does not exhibit a regular rotation profile. The
fit to the central spectrum is improved by the addition of a broad component which
appears to represent a non-flat continuum in this case. Data for this galaxy are shown
in Figure 3.8 (see key in Figure 3.1 for an explanation of these plots).
NGC 2892 (UGC 5073) This elliptical galaxy has no apparent nuclear dust. It
has a radio core on VLBA scales and a twin-jet morphology on VLA scales. The
STIS slits were aligned approximately parallel to the galaxy major axis, however a
certain degree of freedom was allowed in slit placement to allow reasonable observing

91. 73
windows. The central kinematic and flux properties are listed in Table 3.12; the gas
does not exhibit a regular rotation profile. The fit to the central spectrum is not
significantly improved by the addition of a broad component. Data for this galaxy
are shown in Figure 3.9 (see key in Figure 3.1 for an explanation of these plots).
NGC 3801 (UGC 6635) This S0/a galaxy has a complex nuclear dust morphol-
ogy with a large scale dust lane (width : length = 0.12). It has a twin-jet radio
morphology on VLA scales. The STIS slits were aligned parallel to the galaxy major
axis. The central kinematic and flux properties are listed in Table 3.13; the gas does
not exhibit a regular rotation profile. The fit to the central spectrum is not signif-
icantly improved by the addition of a broad component. Data for this galaxy are
shown in Figure 3.10 (see key in Figure 3.1 for an explanation of these plots).
NGC 3862 (UGC 6723) This elliptical galaxy has a nuclear dust disk (b/a ∼
0.99). It has a core-jet radio morphology on VLBA and VLA scales. The STIS slits
were aligned approximately perpendicular to the radio jet as the nuclear isophotal
position angles are poorly constrained. The central kinematic and flux properties are
listed in Table 3.14; the gas does not exhibit a regular rotation profile. The fit to the
central spectrum is improved by the addition of a broad component which appears to

92. 74
represent a non-flat continuum in this case. Data for this galaxy are shown in Figure
3.11 (see key in Figure 3.1 for an explanation of these plots).
UGC 7115 This elliptical galaxy has a nuclear dust disk (b/a ∼ 0.95). It has a
core-jet radio morphology on VLA scales. The STIS slit were aligned approximately
perpendicular to the radio jet as the nuclear isophotal position angles are poorly con-
strained, a certain degree of freedom was allowed in slit placement to allow reasonable
observing windows. This galaxy was observed in only one slit position, as we also
required WFPC2 observations of this target in order to measure the central photo-
metric properties (see Verdoes Kleijn et al., 2002a). The central kinematic and flux
properties are listed in Table 3.15; the gas exhibits a regular rotation profile. The
fit to the central spectrum is not significantly improved by the addition of a broad
component. Data for this galaxy are shown in Figure 3.12 (see key in Figure 3.1 for
an explanation of these plots).
NGC 4261 (UGC 7360) This E2-3 galaxy has a nuclear dust disk (b/a = 0.46).
It has a twin-jet radio morphology on VLBA and VLA scales. The STIS slits were
aligned parallel to the galaxy major axis. The central kinematic and flux properties
are listed in Table 3.16. The nucleus of this galaxy lies closer to one of the side slits

93. 75
(slit one) than the central position, however it is still possible to see a clear rotation
curve along that slit. The fit to the central spectrum is improved by the addition of
a broad component. Data for this galaxy are shown in Figure 3.13 (see key in Figure
3.1 for an explanation of these plots).
NGC 4335 (UGC 7455) This elliptical galaxy has a nuclear dust disk (b/a =
0.41). It has a radio core on VLBA scales and a twin-jet morphology on VLA scales.
The STIS slits were aligned parallel to the galaxy major axis. The central kinematic
and flux properties are listed in Table 3.17; the gas exhibits a regular rotation profile.
In the positive offset side slit there is a dip in the velocity dispersion profile at the
position closest to the nucleus. See also Verdoes Kleijn et al. (2002b). The fit to the
central spectrum is improved by the addition of a broad component. Data for this
galaxy are shown in Figure 3.14 (see key in Figure 3.1 for an explanation of these
plots).
NGC 4374 (M84; UGC 7494) This E1 galaxy has a nuclear dust lane (width :
length = 0.15). It has a core-jet radio morphology on VLBA scales, and a twin-jet
morphology on VLA scales. The STIS slits were aligned approximately perpendicular
to the radio jets, which lies close to the major axis of the emission line gas. The

94. 76
central kinematic and flux properties are listed in Table 3.18; the gas exhibits a
regular rotation profile. See also Bower et al. (1998). The fit to the central spectrum
is not significantly improved the addition of a broad component. Data for this galaxy
are shown in Figure 3.15 (see key in Figure 3.1 for an explanation of these plots).
NGC 4486 (M87; UGC 7654) This elliptical galaxy has an irregular nuclear dust
morphology. It has a core-jet radio morphology on VLBA and VLA scales. The STIS
slits were aligned to trace morphological features in the emission line gas across the
nuclear region of this galaxy. The central kinematic and flux properties are listed in
Table 3.19; the gas exhibits a regular rotation profile. The fit to the central spectrum
is improved by the addition of a broad component. Data for this galaxy are shown
in Figure 3.16 (see key in Figure 3.1 for an explanation of these plots).
NGC 5127 (UGC 8419) This elliptical peculiar galaxy has a nuclear dust lane
(width : length = 0.25). It has a radio core on VLBA scales and a twin-jet morphol-
ogy on VLA scales. The STIS slits were aligned parallel to the galaxy major axis.
The central kinematic and flux properties are listed in Table 3.20; the gas exhibits a
regular rotation profile. The fit to the central spectrum is not significantly improved
by the addition of a broad component. Data for this galaxy are shown in Figure 3.17

95. 77
(see key in Figure 3.1 for an explanation of these plots).
NGC 5141 (UGC 8433) This S0 galaxy has a nuclear dust lane (width : length =
0.25). It has a core-jet radio morphology on VLBA scales and a twin-jet morphology
on VLA scales. The STIS slits were aligned parallel to the galaxy major axis. The
central kinematic and flux properties are listed in Table 3.21; the gas exhibits a
regular rotation profile. The fit to the central spectrum is improved by the addition
of a broad component. Data for this galaxy are shown in Figure 3.18 (see key in
Figure 3.1 for an explanation of these plots).
NGC 5490 (UGC 9058) This elliptical galaxy has a nuclear dust lane (width :
length = 0.35). It has a core-jet radio morphology on VLBA scales and a twin-jet
morphology on VLA scales. The STIS slits were aligned parallel to the galaxy major
axis. The central kinematic and flux properties are listed in Table 3.22; the gas does
not exhibit a regular rotation profile. The fit to the central spectrum is improved by
the addition of a broad component. Data for this galaxy are shown in Figure 3.19
(see key in Figure 3.1 for an explanation of these plots).

96. 78
NGC 7052 (UGC 11718) This elliptical galaxy has a nuclear dust disk (b/a =
0.30). It has a twin-jet radio morphology on VLBA scales and a core-jet morphology
on VLA scales. The STIS slits were aligned parallel to the galaxy major axis. The
central kinematic and flux properties are listed in Table 3.23; the gas exhibits a regular
rotation profile. The fit to the central spectrum is not significantly improved by the
addition of a broad component. Data for this galaxy are shown in Figure 3.20 (see
key in Figure 3.1 for an explanation of these plots).
UGC 12064 This S0 galaxy has a nuclear dust disk (b/a = 0.54). It has a twin-jet
radio morphology on VLA scales. The STIS slits were aligned parallel to the dust
disk major axis. The central kinematic and flux properties are listed in Table 3.24;
the gas exhibits a regular rotation profile. The fit to the central spectrum is improved
by the addition of a broad component. Data for this galaxy are shown in Figure 3.21
(see key in Figure 3.1 for an explanation of these plots).
NGC 7626 (UGC 12531) This elliptical peculiar galaxy has a nuclear dust lane
(width : length = 0.17). It has a core-jet radio morphology on VLBA scales and
a twin-jet morphology on VLA scales. The STIS slits were aligned parallel to the
galaxy major axis. The central kinematic and flux properties are listed in Table 3.25;

97. 79
the gas exhibits a regular rotation profile. The fit to the central spectrum is improved
by the addition of a broad component. Data for this galaxy are shown in Figure 3.22
(see key in Figure 3.1 for an explanation of these plots).
3.5 Interpretation and Discussion
In our initial interpretation we have focussed on understanding the general parameters
of the data set. We will undertake more detailed analyses in future work that we
outline in §6 below. Here, we first describe the categorization of sources as rotating
and non-rotating systems based on the observed kinematics (§5.1). We then discuss
the ionization states of the nuclear regions (§5.2). We go on to discuss the presence
of broad components in these nuclei and a more detailed analysis of the line shapes
(§5.3).
3.5.1 Rotators and non-rotators
By inspecting maps of the central kinematics and the velocity profiles along each slit
(as presented above in figures 3.2 to 3.22), we have classified, by eye, the galaxies
into two classes: rotators and non-rotators. Rotators are systems where we see pat-
terns reminiscent of rotation curves; in non-rotators we find no such patterns - the

98. 80
kinematics seem either irregular or organized in some manner that does not represent
regular rotation. We do not include NGC 741 in discussions of kinematics as very
few points were well fit during our analysis. We classify 67% (N = 14/21) of the
UGC FR-I galaxies as rotators. 73% of galaxies with dust disks (N = 8/11), 100%
of galaxies with dust lanes (N = 5/5) and 50% of galaxies with complex dust or no
dust (N = 2/4) are rotators.
We have made use of the mean velocity dispersion
σ100pc =
1
N i
σi : xi ≤ 100pc, (3.4)
and the difference in mean velocities on each side of the nucleus
∆100pc =
1
N1 i
vi : −100pc ≤ xi < 0 −


1
N2 j
vj : 0 < xj ≤ 100pc

 (3.5)
within 100 pc of the brightest pixel as illustrative of the global kinematic pa-
rameters along the central slit2
. These parameters are shown in Table 3.30 for each
galaxy, along with the mean properties for each class of galaxy. In Figure 3.23 we
2
For NGC 4261 we used the offset slit closest to the nucleus as explained above.