0807.3114/ms.tex
1: %% \documentclass[12pt,preprint]{aastex}
2: %% \documentclass[manuscript]{aastex}
3: %% \documentclass[preprint2]{aastex}
4: %% \documentclass[preprint2,longabstract]{aastex}
5: \documentclass{emulateapj}
6: 
7: \slugcomment{To appear in the {\em Astrophysical Journal}}
8: \shorttitle{Low-Mass Active Galactic Nucleus GH\,10}
9: \shortauthors{Wrobel et al.}
10: \begin{document}
11: \title{Steep-Spectrum Radio Emission from the Low-Mass 
12:        Active Galactic Nucleus GH\,10}
13: \author{J. M. Wrobel\altaffilmark{1},
14:         J. E. Greene\altaffilmark{2},
15:         L. C. Ho\altaffilmark{3},
16:     and J. S. Ulvestad\altaffilmark{1}}
17: 
18: \altaffiltext{1}{National Radio Astronomy Observatory, P.O. Box O,
19: Socorro, NM 87801; jwrobel@nrao.edu, julvesta@nrao.edu}
20: \altaffiltext{2}{Department of Astrophysical Sciences, Princeton
21: University, Princeton, NJ 08544; jgreene@astro.princeton.edu}
22: \altaffiltext{3}{The Observatories of the Carnegie Institution of
23: Washington, 813 Santa Barbara Street, Pasadena, CA 91101;
24: lho@ociw.edu}
25: 
26: \begin{abstract}
27: GH\,10 is a broad-lined active galactic nucleus (AGN) energized by a
28: black hole of mass 800,000~$M_\odot$.  It was the only object detected
29: by Greene et al.\ in their Very Large Array (VLA) survey of 19
30: low-mass AGNs discovered by Greene \& Ho.  New VLA imaging at 1.4,
31: 4.9, and 8.5~GHz reveals that GH\,10's emission has an extent of less
32: than 320~pc, has an optically-thin synchrotron spectrum with a
33: spectral index $\alpha = -0.76\pm0.05$ ($S_\nu \propto
34: \nu^{+\alpha}$), is less than 11\% linearly polarized, and is steady -
35: although poorly sampled - on timescales of weeks and years.
36: Circumnuclear star formation cannot dominate the radio emission,
37: because the high inferred star formation rate, 18~$M_\odot$~yr$^{-1}$,
38: is inconsistent with the rate of less than 2~$M_\odot$~yr$^{-1}$
39: derived from narrow H$\alpha$ and [O~II] $\lambda$3727 emission.
40: Instead, the radio emission must be mainly energized by the low-mass
41: black hole.  GH\,10's radio properties match those of the
42: steep-spectrum cores of Palomar Seyfert galaxies, suggesting that,
43: like those Seyferts, the emission is outflow-driven.  Because GH\,10
44: is radiating close to its Eddington limit, it may be a local analog of
45: the starting conditions, or seeds, for supermassive black holes.
46: Future imaging of GH\,10 at higher linear resolution thus offers an
47: opportunity to study the relative roles of radiative versus kinetic
48: feedback during black-hole growth.
49: \end{abstract}
50: 
51: \keywords{galaxies: active ---
52:           galaxies: individual (GH\,10) ---
53:           galaxies: nuclei ---
54:           galaxies: Seyfert ---
55:           radio continuum: galaxies}
56: 
57: \section{Motivation}
58: 
59: Although little is known about the mass function of nuclear black
60: holes with mass below about $10^6~M_\odot$, this is an important
61: regime for several reasons.  For example, the remarkable correlation
62: between host bulge velocity dispersion and black hole mass strongly
63: hints that black holes play an essential role in galaxy evolution
64: \citep{kor04}, yet there are few observational constraints on the
65: starting conditions, or so-called seeds, for the supermassive black
66: holes with masses above about $10^6~M_\odot$.  Also, until the {\em
67: Laser Interferometer Space Antenna\/} launches, an understanding of
68: the low-mass end of the black-hole mass function may provide one of
69: the few constraints on these seed black holes \citep{hug02}.
70: Moreover, it remains to be established whether small galaxies can
71: commonly host nuclear black holes and, if so, whether the same mass
72: versus dispersion relation mentioned above is obeyed.
73: 
74: Motivated by these and similar issues, \citet{gre04} conducted the
75: first systematic search for a population of black holes with mass
76: below about $10^6~M_\odot$.  They used SDSS DR1 to select a sample of
77: broad-lined active galactic nuclei (AGNs) with virial mass estimates
78: $10^{5-6}~M_\odot$.  This sample of 19 objects was the first uniformly
79: selected sample of low-mass AGNs.  The sample is characterized by
80: Eddington ratios, $L_{\rm bol}/L_{\rm Edd}$, that are close to unity,
81: implying radiatively efficient accretion onto these black holes.  With
82: its low-mass black holes and high Eddington ratios, this sample probes
83: a poorly explored region of parameter space, potentially providing new
84: insights into the physical drivers of radio properties in AGNs.  For
85: this reason, \citet{gre06} observed the sample at 4.9 GHz with the
86: Very Large Array (VLA).  A single object, \object{GH 10}, was detected
87: with a 4.9-GHz--to--optical flux ratio $R \sim 2.8$, formally
88: radio-quiet \citep{kel89}.  A stacked image of the remaining 18
89: objects yielded an upper limit of $R \le 0.27$, emphasizing the
90: atypical traits of GH\,10.
91: 
92: To help understand the physical origin of these atypical traits, new
93: multi-frequency VLA data were obtained on GH\,10 and are presented in
94: Section 2.  As summarized in Section 3.1, the radio emssion from
95: GH\,10 is found to be compact, steep-spectrum, unpolarized, and steady
96: in time.  A star-formation origin for this emission is considered, and
97: dismissed, in Section 3.2, while Section 3.3 explores the implications
98: of a black-hole origin for the radio emission.  The paper ends, in
99: Section 4, with a summary of the key findings and suggestions for
100: future directions.
101: 
102: The following cosmological parameters are assumed: $H_0 =
103: 71$~km~s$^{-1}$~Mpc$^{-1}$, $\Omega_m = 0.27$, and $\Omega_\Lambda =
104: 0.75$ \citep{spe03}, leading to a luminosity distance of 363~Mpc
105: \citep{gre04,gre06}.  Also, the sign convention adopted for the
106: spectral index $\alpha$ near a frequency $\nu$ is the flux density
107: $S_\nu \propto \nu^{+\alpha}$.
108: 
109: \section{VLA Imaging}
110: 
111: The VLA was used in the A configuration \citep{tho80} to observe
112: GH\,10 under proposal code AG745 on UT 2007 Jun 18 and 2007 Jul 15.
113: Observations were phase-referenced to the calibrator J1224+0330 whose
114: positional accuracy was 0.01\arcsec.  The switching angle was
115: 5\arcdeg, while the switching times were 560~s, 260~s, and 200~s for
116: observations at center frequencies of 1.4000, 4.8601, and 8.4601~GHz,
117: respectively.  These frequencies will be abbreviated as 1.4, 4.9, and
118: 8.5~GHz hereafter.  The {\em a priori\/} pointing position for GH\,10
119: was centered 3\arcsec\, south of the SDSS position \citep{gre04} to
120: avoid any phase-center artifacts.  Observations were made assuming a
121: coordinate equinox of 2000.  Data were acquired with a bandwidth of
122: 100~MHz for each circular polarization.  Observations of 3C\,286 were
123: used to set the amplitude scale to an accuracy of about 3\%.
124: 
125: On 2007 Jun 18, net exposure times were 1500~s, 7800~s, and 8300~s at
126: 1.4, 4.9, and 8.5~GHz, respectively.  Twenty-five to 27 antennas
127: provided data of acceptable quality, with most of the data loss
128: attributable to EVLA retrofitting activities.  The observations at
129: 1.4~GHz were made at a large hour angle and were followed by
130: alternating observations at 4.9 and 8.5~GHz to sample similar hour
131: angles.  On 2007 Jul 15, 1.4-GHz observations were made at transit and
132: resulted in a net exposure time of 2800~s.  Twenty-four antennas
133: provided data of acceptable quality, with most of the data loss
134: attributable to EVLA retrofitting activities and software failures.
135: The data were calibrated using the 2007 Dec 31 release of the NRAO
136: AIPS software.  No self-calibrations were performed.  No polarization
137: calibration was performed, as only upper limits to the linear
138: polarization percentages were sought.  The AIPS task {\tt imagr} was
139: used to form and deconvolve images of the emission from GH\,10.
140: 
141: Following \citet{wro00}, {\tt imagr} was used to form and deconvolve
142: Stokes $I\/$ images at 4.9 and 8.5~GHz at a matched resolution at FWHM
143: of 0.42\arcsec.  This led to detections of a compact source at both
144: frequencies, with no evidence for adjacent emission on larger scales.
145: The 4.9~GHz image appears in Figure~1.  Quadratic fits in the image
146: plane yielded the integrated flux densities and the two-point spectral
147: index given in Table~1.  These flux densities from 2007 Jun 18 are
148: plotted in Figure~2.  The image fits also yielded identical
149: deconvolved diameters of less than 0.21\arcsec, as well as almost
150: identical positions.  The 4.9-GHz position is $\alpha(J2000) = 12^{h}
151: 40^{m} 35^{s}.825$ and $\delta(J2000) = -00^{\circ} 29' 19''.53$, with
152: a 1-D error of 0.1\arcsec, the quadratic sum of terms due to the
153: signal-to-noise ratio (SNR) of the detection (less than 0.01\arcsec),
154: the phase-calibrator position error (0.01\arcsec), and the
155: phase-referencing strategies (estimated to be 0.1\arcsec).  This
156: 4.9-GHz position is consistent with the SDSS position \citep{gre04}
157: marked in Figure~1.  {\tt imagr} was also used to form
158: matched-resolution images of Stokes $Q\/$ and $U\/$ at 4.9 and
159: 8.5~GHz.  Those images led to no detections, with the most
160: constraining upper limit to the linear polarization percentage being
161: less than 11\% at 4.9~GHz.
162: 
163: Following \citet{ho01b}, {\tt imagr} was used in a two-step approach
164: to form and deconvolve Stokes $I\/$ images at 1.4~GHz.  First, a
165: search for confusing sources was done by using the inner array
166: baselines to image, at low resolution, the primary antenna beam to the
167: half-power point.  Second, all array baselines were used to image
168: GH\,10 and its confusing sources, simultaneously and at a high
169: resolution of about 1.4\arcsec\, at FWHM.  From the GH\,10 data
170: acquired at transit on 2007 Jul 15, this approach led to a
171: high-resolution image mildly affected by confusion, showing a compact
172: detection plus no evidence for adjacent emission on larger scales; a
173: quadratic fit in the image plane yielded the integrated flux density
174: in Table~1 and a deconvolved diameter of less than 0.72\arcsec.  From
175: the GH\,10 data acquired at a large hour angle on 2007 Jun 18, the
176: high-resolution image was more degraded by confusion so only a
177: parabolic fit in the image plane was made.  This yielded the peak flux
178: density in Table~1 that is plotted in Figure~2.  That figure also
179: shows the spectral index $\alpha = -0.76\pm0.05$ derived from all the
180: photometry on 2007 Jun 18.
181: 
182: Finally, the 4.9-GHz data were used to produce an image at the FWHM
183: resolution of the 1.4-GHz data on 2007 Jun 18.  A parabolic fit to
184: that 4.9-GHz image gave the flux density listed in Table~1, leading in
185: turn to the tabulated spectral index between 1.4 and 4.9~GHz.  Also,
186: that 4.9-GHz flux density is consistent with the value obtained at a
187: FWHM resolution of 0.42\arcsec\, implying no additional emission on
188: scales between 0.42\arcsec and about 1.4\arcsec.
189: 
190: \section{Implications}
191: 
192: \subsection{Radio Properties}
193: 
194: The new VLA data imply that the emission from GH\,10 is compact, with
195: a diameter of less than 0.21\arcsec\, (320~pc) at 4.9~GHz (Fig.~1) and
196: 8.5~GHz, and less than 0.72\arcsec\, (1.1~kpc) at 1.4~GHz.  These
197: upper limits are derived from data with SNRs of 15 or more.  The new
198: 1.4-GHz images show only a compact component and do not confirm the
199: prior hint from low-SNR data of marginally resolved emission with a
200: FWHM diameter of 5\arcsec\, \citep[7.5~kpc;][]{whi97}.  Similarly, the
201: new 4.9-GHz image does not confirm the prior suggestion from low
202: SNR-data of marginally-resolved emission with a FWHM diameter of
203: 0.3\arcsec\, \citep[530~pc;][]{gre06}.
204: 
205: The emission from GH\,10 has a steep radio spectrum, with a
206: three-point spectral index of $\alpha = -0.76\pm0.05$ measured on 2007
207: Jun 18 and thus free from concerns about time variability (Fig.~2).
208: This spectral index agrees with the two-point, matched-resolution
209: spectral indices given in Table~1, as is expected for compact,
210: unresolved emission.  The single-epoch radio spectrum shown for GH\,10
211: in Figure~2 - the first ever measured for any of the 19 low-mass AGNs
212: discovered by Greene and Ho \citep{gre04,gre06} - is consistent with
213: optically thin synchrotron emission.  This emission is less than 11\%
214: linearly polarized at 4.9~GHz, well below the 73\% expected for
215: electrons with a power-law energy distribution, with index $p = 1 -
216: 2\alpha$, in a uniform magnetic field \citep[][page 197]{shu91}.
217: Faraday depth effects could also be at play.
218: 
219: Variability timescales for AGNs are expected to scale with black-hole
220: mass \citep[e.g., ][]{ede99}, and have not yet been explored for any
221: of the 19 low-mass Greene and Ho objects.  The new VLA photometry can
222: be used to assess GH\,10's time variabilty, for future comparison with
223: other GH\,10 timescales to search for physical linkages.  Using only
224: A-configuration data, the ratio of the 4.9-GHz flux density on 2007
225: Jun 18 (0.75$\pm$0.05~mJy) to that on 2004 Oct 9
226: \citep[0.7$\pm$0.1~mJy;][]{gre06} is 7$\pm$16\%, consistent with
227: steady emission on timescales of years.  The ratio of the 1.4-GHz flux
228: density on 2007 Jul 15 (1.88$\pm$0.13~mJy) to that on 2007 Jun 18
229: (1.80$\pm$0.09~mJy) is 4$\pm$9\%, also consistent with steady emission
230: on timescales of weeks.  Also, folding in the B-configuration survey
231: of \citet{whi97}, the 1.4-GHz flux density of GH\,10 on 1998 Aug 11
232: was 1.16$\pm$0.16 mJy (peak) and 1.29$\pm$0.16 mJy (integrated).  For
233: such a low-SNR and compact source, \citet{dev04} advocate using the
234: peak flux density and seeking variability at the 4 $\sigma$, or
235: greater, level.  For GH\,10, the ratio of the flux density on 2007 Jul
236: 15 (1.88$\pm$0.13 mJy) to that earlier peak (1.16$\pm$0.16 mJy) is
237: 62$\pm$16\%, a change of less than 4 $\sigma$ on a timescale of almost
238: a decade.
239: 
240: In summary, the radio emssion from GH\,10 is compact, steep-spectrum,
241: unpolarized, and steady - although sparsely sampled - on timescales of
242: weeks and years.  The integrated 1.4-GHz flux density corresponds to a
243: power of $3.0 \times 10^{22}$~W~Hz$^{-1}$, while that at 4.9~GHz
244: corresponds to a power of $1.2 \times 10^{22}$~W~Hz$^{-1}$.
245: 
246: \subsection{Radio Emission from Star Formation?}
247: 
248: The radio emission from GH\,10 might be driven, in part, by
249: circumnuclear star formation.  Applying {\tt scanpi} to {\em Infrared
250: Astronomical Satellite\/} data at 60~$\mu$m, GH\,10 is undetected with
251: a 3 $\sigma$ upper limit of 0.15~Jy.  Equation (2) of \citet{yun01}
252: yields an upper limit to the 60-$\mu$m power.  Inserting that limit
253: into the radio-infrared relation in equation (4) of \citet{yun01} sets
254: an upper limit of $< 1.9 \times 10^{22}$~W~Hz$^{-1}$ to the 1.4-GHz
255: power expected from any star formation.  This limit is below the
256: observed power of $3.0 \times 10^{22}$~W~Hz$^{-1}$, but not by much.
257: 
258: A more telling analysis follows from consideration of the host
259: galaxy's properties, especially as they relate to estimates for a
260: star-formation rate.  Keck spectroscopy, with a 1\arcsec\, (1.5~kpc)
261: aperture, exhibit the Mg I{\it b} and Ca II infrared triplets
262: characteristic of an old stellar population \citep{bar05}.  {\em
263: Hubble Space Telescope\/} $B$ and $I$ imaging, with a resolution at
264: FWHM of 0.12\arcsec\, (170~pc), reveals a smooth, spheroidal-looking
265: system with a steep Sersic index $n=4$ and a color $B-I=2$ that is
266: characteristic of Sa or spheroidal galaxies \citep{gre08}.  SDSS
267: spectroscopy with a 3\arcsec\, (4.5~kpc) aperture shows narrow
268: H$\alpha$ and [O~II] $\lambda$3727 lines but, as a Balmer decrement
269: could not be robustly measured, an estimate of the internal reddening
270: is unavailable for GH\,10 \citep{gre04}.  However, if the average
271: reddening advocated for the 19 Greene and Ho objects by \citet{gre08}
272: is adopted and the reddening curve of \citet{car89} is applied, then
273: the correction factors for narrow H$\alpha$ and [O~II] $\lambda$3727
274: are 1.3 and 1.7, respectively.  Then using equations (2)-(3) of
275: \citet{ken98}, those corrected line luminosities for GH\,10 each imply
276: a star formation rate no higher than 2~$M_\odot$~yr$^{-1}$.  These
277: rate estimates are, by construction, upper limits because they neglect
278: the portion of the line-emitting gas that is energized by the AGN
279: rather than by star formation.  Moreover, in order for GH\,10 to be
280: within the 2 $\sigma$ range of the radio-infrared relation mentioned
281: above, the reddening at H$\alpha$ would need to be 2 mag, far more
282: than the average (0.3 mag) and largest (0.5 mag) reddening advocated
283: for the Greene and Ho objects \citep{gre08}.  Fortunately, GH\,10 is
284: also detected in a {\em Spitzer\/} IRS spectrum, boding well for a
285: future estimate of a star formation rate that is less affected by
286: reddening concerns (C. E. Thornton et al., in preparation).
287: 
288: For comparison, using equation (13) of \citet{yun01}, the observed
289: 1.4-GHz power of GH\,10 would correspond to a star-formation rate of
290: 18~$M_\odot$~yr$^{-1}$.  Such a high rate is clearly inconsistent with
291: the line-based upper limit of 2~$M_\odot$~yr$^{-1}$ derived above.
292: Circumnuclear star formation cannot, therefore, dominate the radio
293: emission from GH\,10.
294: 
295: \subsection{Radio Emission from the AGN}
296: 
297: Based on the above analysis, it seems plausible that the radio
298: emission from GH\,10 is mainly energized by its low-mass black hole.
299: GH\,10 is radiating at about twice its Eddingtion luminosity, with
300: this estimate being uncertain by a factor of 3 \citep{gre04}.  Thus
301: comparisons of GH\,10 with other broad-lined AGNs with near-unity
302: Eddington ratios are especially apt.  The 4.9-GHz--to--optical flux
303: ratio for GH\,10 implies $R \sim 2.8$, so GH\,10 is radio-quiet
304: \citep{kel89}, a trait shared by most quasars at $z < 0.5$
305: \citep{ho01a,gre04,gre06}.  There is also a faint, compact {\em
306: Chandra\/} detection within 1\arcsec\, (1.5~kpc) of the SDSS position
307: of GH\,10, and its spectral properties resemble those of high-mass
308: black holes \citep{gre07a}.  The ratio of the 4.9-GHz luminosity, $5.9
309: \times 10^{38}$~ergs~s$^{-1}$, to the hard X-ray luminosity, $1.9
310: \times 10^{42}$~ergs~s$^{-1}$, implies $R_{\rm X} = 10^{-3.5}$
311: \citep{ter03}.  Such a value is also consistent with the values shown
312: by most $z < 0.5$ quasars and by most luminous Seyfert 1 galaxies
313: \citep{ho08}.  Unfortunately, such systems are too faint and too
314: distant to be resolved on sub-kiloparsec scales in VLA images, so no
315: structural comparisons can be made.
316: 
317: GH\,10's values for $R\/$ and $R_{\rm X}$, plus its radio and hard
318: X-ray luminosities, also resemble those for the more luminous Palomar
319: Seyfert galaxies \citep{pan07}.  VLA imaging at 1.4 and 4.9~GHz of the
320: Palomar Seyfert galaxies, with a resolution of about 1\arcsec\,
321: (100~pc at the median distance of 20~Mpc), shows a wide range of radio
322: powers ($10^{18}-10^{25}$~W~Hz$^{-1}$), spectral indices (+0.5 to -1),
323: and linear sizes (a few tens of parsecs to 15~kpc)
324: \citep{ho01b,ulv01}.  Their structures mainly show a compact core,
325: either unresolved or slightly resolved, sometimes with adjacent
326: elongated, jetlike/outflow components.  Linearly polarized emission,
327: while rarely detected, is preferentially associated with features
328: adjacent to the cores.  Thus GH\,10's radio properties most resemble
329: those of the steep-spectrum cores of the Palomar Seyferts, for which
330: the emission is unmodified by the effects of internal and/or external
331: opacity.  Like the Palomar Seyferts, the steep-spectrum emission from
332: GH\,10 could be outflow-driven.  Moreover, the lower escape velocity
333: from GH\,10's dwarf host galaxy \citep{bar05} would make it easier for
334: any synchrotron-emitting plasma to leave the vicinity of the AGN,
335: thereby favoring a steep, rather than flat or inverted, radio
336: spectrum.
337: 
338: This work has established that the radio emission from GH\,10 is
339: compact, steep-spectrum, unpolarized, and steady.  Still, GH\,10 is
340: known to be atypical when compared to the other GH objects, which are
341: especially radio-quiet with $R \le 0.27$ from a stacking analysis
342: \citep{gre06}.  Following those authors, by analogy with supermassive
343: and stellar-mass black holes, a near-unity Eddington ratio suggests
344: that the low-mass black hole GH\,10 is presently in the very high
345: state, with its entry into that state from the high/soft state
346: accompanied by an optically thin ejection event in the radio.  Within
347: this framework, and barring opacity effects, the radio emission from
348: GH\,10 would be the slowly fading, steep-spectrum remnant from that
349: ejection event, whose power corresponds to the quenched radio levels
350: expected for the high/soft state \citep[e.g.,][]{mac03}.  Also, a
351: radio detection of only one of the 19 GH objects suggests, at face
352: value, that such radio ejections from GH objects occur only about 5\%
353: of the time.  Interestingly, a 5\% duty cyle resembles that
354: established observationally for radio flares from microquasars
355: \citep[e.g.,][]{nip05}.
356: 
357: Because the low-mass black hole GH\,10 is radiating close to its
358: Eddington limit, it may be a local analog of the starting conditions,
359: or seeds, for supermassive black holes.  From the present work, the
360: diameter of GH\,10's radio emission must be less than 0.21\arcsec\,
361: (320~pc).  If GH\,10 is indeed driving some of this emission, then
362: 1.4-GHz imaging with Very Long Baseline Interferometry (VLBI) at a
363: resolution of 0.005\arcsec\, (7.5~pc) will reveal the physical
364: connection between the black hole and the synchrotron-emitting plasma.
365: Although GH\,10 is a faint radio source (Table~1), VLBI imaging of
366: such sources is feasible \citep[e.g.,][]{wro06} and would offer a
367: means to study the relative roles of radiative versus kinetic feedback
368: during black-hole growth \citep[e.g.,][]{dim05,kuh05}.
369: 
370: \section{Summary and Future Directions}
371: 
372: GH\,10 was the only object detected by \citet{gre06} in their VLA
373: survey of the 19 low-mass AGNs discovered by \citet{gre04}.  New VLA
374: imaging reveals that the emssion from GH\,10 has an extent of less
375: than 320~pc, with an optically thin synchrotron spectrum that is less
376: than 11\% linearly polarized.  The radio emission is steady, although
377: poorly sampled, on timescales of weeks and years.  Circumnuclear star
378: formation cannot dominate the radio emission, because the inferred
379: star formation rate, 18~$M_\odot$~yr$^{-1}$, is inconsistent with the
380: upper limit of 2~$M_\odot$~yr$^{-1}$ derived from the narrow optical
381: emission lines.  Instead, the radio emission must be dominantly
382: energized by the low-mass black hole, which is thought to be radiating
383: near its Eddington limit.
384: 
385: The radio properties of GH\,10 strongly resemble those of the
386: steep-spectrum cores of Palomar Seyfert galaxies, which feature
387: outflow-driven emission unmodified by the effects of internal and/or
388: external opacity.  This suggest that, like those Palomar Seyferts, the
389: radio emission from GH\,10 is outflow-driven, a suggestion that can be
390: tested with future VLBI imaging.  Such imaging could also offer an
391: opportunity to study the relative roles of radiative versus kinetic
392: feedback during black-hole growth, given that GH\,10 may be a local
393: analog of the starting conditions, or seeds, for supermassive black
394: holes.  Finally, the compact, steep-spectrum, unpolarized, and steady
395: nature of the emission from GH\,10 will guide future observations of
396: the new 1.4-GHz detections of low-mass AGNs reported by
397: \citet{gre07b}.
398: 
399: %% Included in this acknowledgments section are examples of the
400: %% AASTeX hypertext markup commands. Use \url without the optional [HREF]
401: %% argument when you want to print the url directly in the text. Otherwise,
402: %% use either \url or \anchor, with the HREF as the first argument and the
403: %% text to be printed in the second.
404: 
405: \acknowledgments We acknowledge useful feedback from an anonymous referee.
406: 
407: {\it Facilities:} \facility{VLA}
408: 
409: \begin{thebibliography}{}
410: 
411: \bibitem[Barth et al.(2005)]{bar05} Barth, A. J., Greene, J. E., \&
412:  Ho, L. C. 2005, \apj, 619, L151
413: \bibitem[Cardelli et al.(1989)]{car89} Cardelli, J. A., Clayton,
414:  G. C., \& Mathis, J. S. 1989, \apj, 345, 245
415: \bibitem[de Vries et al.(2004)]{dev04} de Vries, W. H., Becker, R. H.,
416:  White, R. L., \& Helfand, D. J. 2004, \aj, 127, 2565
417: \bibitem[Di Matteo et al.(2005)]{dim05} Di Matteo, T., Springel, V.,
418:  \& Hernquist, L. 2005, \nat, 433, 604
419: \bibitem[Edelson \& Nandra(1999)]{ede99} Edelson, R., \& Nandra, K. 
420:  1999, \apj, 514, 682
421: \bibitem[Greene \& Ho(2004)]{gre04} Greene, J. E., \& Ho, L. C.  2004,
422:  \apj, 610, 722
423: \bibitem[Greene \& Ho(2007a)]{gre07a} Greene, J. E., \& Ho, L. C.
424:  2007a, \apj, 656, 84
425: \bibitem[Greene \& Ho(2007b)]{gre07b} Greene, J. E., \& Ho, L. C.
426:  2007b, \apj, 670, 92
427: \bibitem[Greene et al.(2006)]{gre06} Greene, J. E., Ho, L. C., \&
428:  Ulvestad, J. S. 2006, \apj, 636, 56
429: \bibitem[Greene et al.(2008)]{gre08} Greene, J. E., Ho, L. C., \&
430:  Barth, A. J. 2008, \apj, submitted
431: \bibitem[Ho(2008)]{ho08} Ho, L. C. 2008, \araa, 46, in press
432:  (astro-ph/0803.2268)
433: \bibitem[Ho \& Peng(2001)]{ho01a} Ho, L. C., \& Peng, C. Y. 2001,
434:  \apj, 555, 650
435: \bibitem[Ho \& Ulvestad(2001)]{ho01b} Ho, L. C., \& Ulvestad, J. S.
436:  2001, \apjs, 133, 77
437: \bibitem[Hughes(2002)]{hug02} Hughes, S. A. 2002, \mnras, 331, 805
438: \bibitem[Kellermann et al.(1989)]{kel89} Kellermann, K. I., Sramek,
439:  R. A., Schmidt, M., Shaffer, D. B., \& Green, R. F. 1989, \aj, 98,
440:  1195
441: \bibitem[Kennicutt(1998)]{ken98} Kennicutt, R. C., Jr. 1998, \araa,
442:  36, 189
443: \bibitem[Kormendy(2004)]{kor04} Kormendy, J. 2004, in Carnegie
444:  Observatories Astrophysics Series, Vol. 1: Coevolution of Black Holes
445:  and Galaxies, ed. L. C. Ho (Cambridge: Cambridge Univ. Press), 1
446: \bibitem[Kuhlen \& Madau(2005)]{kuh05} Kuhlen, M., \& Madau, P. 2005,
447:  \mnras, 363, 1069
448: \bibitem[Maccarone et al.(2003)]{mac03} Maccarone, T. J., Gallo, E.,
449:  \& Fender, R. 2003, \mnras, 345, L19
450: \bibitem[Nipoti et al.(2005)]{nip05} Nipoti, C., Blundell, K. M., \&
451:  Binney, J. 2005, \mnras, 361, 633
452: \bibitem[Panessa et al.(2007)]{pan07} Panessa, F., Barcons, X.,
453:  Bassani, L., Cappi, M., Carrera, F. J., Ho, L. C., \& Pellegrini,
454:  S. 2007, \aap, 467, 519
455: \bibitem[Shu(1991)]{shu91} Shu, F. H. 1991, The Physics of
456:  Astrophysics, Volume 1. Radiation (Mill Valley: University Science
457:  Books)
458: \bibitem[Spergel et al.(2003)]{spe03} Spergel, D. N., et al. 2003,
459:  \apjs, 148, 175
460: \bibitem[Terashima \& Wilson(2003)]{ter03} Terashima, Y., \& Wilson,
461:  A.S. 2003, \apj, 583, 145
462: \bibitem[Thompson et al.(1980)]{tho80} Thompson, A. R., Clark, B. G.,
463:  Wade, C. M., Napier, P. J. 1980, \apjs, 44, 151
464: \bibitem[Ulvestad \& Ho(2001)]{ulv01} Ulvestad, J. S., \& Ho, L. C.
465:  2001, \apj, 558, 561
466: \bibitem[White et al.(1997)]{whi97} White, R. L., Becker, R. H.,
467:  Helfand, D. J., \& Gregg, M. D. 1997, \apj, 475, 479
468:  (http://sundog.stsci.edu/top.html)
469: \bibitem[Wrobel(2000)]{wro00} Wrobel, J. M. 2000, \apj, 531, 716
470: \bibitem[Wrobel \& Ho(2006)]{wro06} Wrobel, J. M., \& Ho, L. C.  2006,
471:  \apj, 646, L95
472: \bibitem[Yun et al.(2001)]{yun01} Yun, M. S., Reddy, N. A., \& Condon,
473:  J. J. 2001, \apj, 554, 803
474: \end{thebibliography}
475: 
476: \clearpage
477: 
478: \begin{deluxetable}{ccclc}
479: \tablecolumns{5}
480: \tablewidth{0pc}
481: \tablecaption{VLA Photometry of GH\,10}\label{tab1}
482: \tablehead{
483: \colhead{}      & \colhead{Resolution} &
484: \colhead{$\nu$} & \colhead{$S_\nu$}    & \colhead{} \\
485: \colhead{Date}  & \colhead{(\arcsec,\arcsec,\arcdeg)} &
486: \colhead{(GHz)} & \colhead{(mJy)}      & \colhead{Fig.~2} \\
487: \colhead{(1)}   & \colhead{(2)}        &
488: \colhead{(3)}   & \colhead{(4)}        & \colhead{(5)}}
489: \startdata
490: 2007 Jun 18& 0.42,0.42,0  & 4.9& 0.75$\pm$0.05\tablenotemark{a}& Yes \\
491:            & 0.42,0.42,0  & 8.5& 0.44$\pm$0.03\tablenotemark{a}& Yes \\
492:            & 1.96,1.44,-47& 1.4& 1.80$\pm$0.09\tablenotemark{b}& Yes \\
493:            & 1.96,1.44,-47& 4.9& 0.72$\pm$0.04\tablenotemark{b}& No  \\
494: 2007 Jul 15& 1.45,1.36,47 & 1.4& 1.88$\pm$0.13                 & No  \\
495: \enddata
496: \tablecomments{
497: Col.~(1): UT observation date.
498: Col.~(2): Elliptical Gaussian resolution at FWHM and its elongation
499: position angle.
500: Col.~(3): Center frequency.
501: Col.~(4): Flux density with error given by the quadratic sum of the
502: 3\% scale error and the 1 $\sigma$ rms of the fit.
503: Col.~(5): Plotted in Figure 2?}
504: \tablenotetext{a}{Two-point spectral index $\alpha = -0.97\pm0.17$.}
505: \tablenotetext{b}{Two-point spectral index $\alpha = -0.73\pm0.06$.}
506: \end{deluxetable}
507: 
508: \clearpage
509: 
510: \begin{figure}
511: \epsscale{1.0}
512: \plotone{f1.eps}
513: \caption{VLA image of Stokes $I\/$ emission from GH\,10 at a frequency
514: of 4.9~GHz and spanning 6\,arcsec\, (9~kpc).  Coordinate labels are
515: relative to the pointing position, which was centered 3\arcsec\, south
516: of the SDSS position to avoid phase-center artifacts.  SDSS position
517: is marked with a cross with a gap.  Natural weighting was used, giving
518: an rms noise of 0.022~mJy~beam$^{-1}$ (1 $\sigma$) and Gaussian beam
519: dimensions at FWHM of 0.42\arcsec\, (hatched circle).  Contours are at
520: $-$6, $-$4, $-$2, 2, 4, 6, 12, and 24 times 1 $\sigma$.  Negative
521: contours are dashed and positive ones are solid.  Linear color scale
522: spans $-$0.1~mJy~beam$^{-1}$ to 0.2~mJy~beam$^{-1}$.}\label{fig1}
523: \end{figure}
524: \clearpage
525: 
526: \begin{figure}
527: \epsscale{1.0}
528: \plotone{f2.eps}
529: \caption{VLA spectrum of GH\,10 as measured at 1.4, 4.9, and 8.5~GHz
530: on 2007 Jun 18.}\label{fig2}
531: \end{figure}
532: 
533: \end{document}
534: