1: \documentclass[12pt,eqsecnum,preprint]{aastex}
2:
3: \begin{document}
4:
5: \title{The FRII Broad Line Seyfert 1 Galaxy: PKSJ 1037-2705}
6: \author{Brian Punsly\altaffilmark{1}, Tracy E. Clarke\altaffilmark{2}, Steven Tingay\altaffilmark{3}}\author{Carlos M.
7: Guti\'{e}rrez\altaffilmark{4}, Jesper Rasmussen\altaffilmark{5} and
8: Ed Colbert\altaffilmark{6}} \altaffiltext{1}{4014 Emerald Street
9: No.116, Torrance CA, USA 90503 and ICRANet, Piazza della Repubblica
10: 10 Pescara 65100, Italy, brian.m.punsly@L-3com.com or
11: brian.punsly@gte.net}\altaffiltext{2}{Naval Research Lab, 4555
12: Overlook Ave SW Washington, DC 20375-5351, Interferometrics Inc.,
13: 13454 Sunrise Valley Drive, Herndon, VA 20171, USA}
14: \altaffiltext{3}{Curtin University of Technology, Department of
15: Imaging and Applied Physics, GPO Box U1987, Perth, Western
16: Australia, 6102, Australia}\altaffiltext{4}{Instituto de Astrofisica
17: de Canarias, Via Lactea, La Laguna, E-38205 Tenerife,
18: Spain}\altaffiltext{5}{Observatories of the Carnegie Institution of
19: Washington, 813 Santa Barbara Street Pasadena, California 91101
20: (Chandra fellow)} \altaffiltext{6}{Physics and Astronomy Department,
21: Johns Hopkins University Baltimore, MD 21218}
22:
23:
24: \begin{abstract}In this article, we demonstrate that PKSJ 1037-2705 has a weak accretion flow
25: luminosity, well below the Seyfert1/QSO dividing line, weak broad
26: emission lines (BELs) and moderately powerful FRII extended radio
27: emission. It is one of the few documented examples of a broad-line
28: object in which the time averaged jet kinetic luminosity,
29: $\overline{Q}$, is larger than the total thermal luminosity (IR to
30: X-ray) of the accretion flow, $L_{bol}$. The blazar nucleus
31: dominates the optical and near ultraviolet emission and is a strong
32: source of hard X-rays. The strong blazar emission indicates that the
33: relativistic radio jet is presently active. The implication is that
34: even weakly accreting AGN can create powerful jets. Kinetically
35: dominated ($\overline{Q}>L_{bol}$) broad-line objects provide
36: important constraints on the relationship between the accretion flow
37: and the jet production mechanism.
38: \end{abstract}
39:
40: \keywords{quasars: general --- individual (PKSJ1037-2705)---
41: galaxies: jets--- galaxies: active--- accretion disks --- black
42: holes}
43:
44: \section{Introduction}It is unclear how the enormous stored energy
45: in the radio lobes of powerful FRII radio sources is related to the
46: thermal luminosity of the accreting gas that flows toward the
47: central black hole. On the one hand, it seems reasonable to expect
48: them to be unrelated, since the bulk of the lobe plasma was ejected
49: from the central engine $\sim 10^{7}$ years earlier than the optical
50: flux from the accretion flow. Therefore, one expects to find many or
51: most FRII sources with fossil lobes, no jets and a weak accretion
52: flow. Yet surprisingly, low frequency surveys such as the 3C survey
53: (which are biased towards sources with strong lobes, i.e. steep
54: spectrum radio emission) contain many sources with FR II lobes in
55: which high dynamic range observations reveal jets that can be traced
56: back to the central quasar, within the resolution of the radio
57: telescope ($\sim$ 1kpc for nearby sources) and many others have
58: strong radio cores and jets on VLBI (parsec) scales (eg, Cygnus A,
59: 3C 175, 3C 334, 3C 215) \citep{bri94,pun01}. On the other hand,
60: there has been considerable research attempting to connect the
61: accretion state with the jet kinetic luminosity, $Q$ - with the
62: conflicting conclusions being the subject of much controversy
63: \citep{raw91,wil99,wan04,bor02}. A truly rigorous and unbiased
64: analysis of the connection between the accretion state and $Q$
65: cannot exclude all of the optically weak FRII radio sources a
66: priori, the narrow line radio galaxies (NLRGs). Many or all of the
67: NLRGs are optically obscured by dusty molecular gas, so a proper
68: treatment must probe inside this gas by looking at the Mid-IR
69: emission. Alternatively, some NLRGs might have weak central AGN and
70: are not significantly obscured. The Spitzer Telescope Mid-IR
71: analysis of a large sample of nearby 3C FRII NLRGs in \citet{ogl06}
72: seemed to indicate that there was a roughly equal mix of FRII NLRGs
73: that had no hidden quasar and those that had a hidden quasar. The
74: implication is that there is a weak central AGN in many of the FRII
75: objects, of which a large proportion have unobstructed lines of
76: sight to the accretion disk. Conversely, Cleary et al 2007 analyzed
77: an even larger sample of Spitzer observations of 3C FRII sources
78: (NLRGS and QSOs) and concluded that orientation effects alone
79: accounted for the differences in FRII NLRGs and FRII QSOs - i.e.,
80: for the most part NLRG had hidden QSOs of the same strength as in
81: FRII QSOs.
82: \par Another closely related debate in the literature is the
83: relationship of accretion luminosity, $L_{bol}$ (which is sometimes
84: expressed in a related form in terms of the Eddington ratio,
85: $R_{Edd}= L_{bol}/L_{Edd}$; where $L_{bol}$ is the bolometric
86: thermal luminosity of the accretion flow and $L_{Edd}$ is the
87: Eddington luminosity) to the jet kinetic luminosity, $Q$. For
88: example, \citet{bor02} concluded that a small $R_{Edd}$ is conducive
89: to strong jets in quasars. The study in \citet{wan04} was an attempt
90: to correlate various properties of the accretion flow with the jet
91: power. They concluded that $Q/L_{bol}$ and $L_{bol}/L_{Edd}$ were
92: inversely correlated in blazars. The inverse correlation claimed
93: between $Q/L_{bol}$ and $L_{bol}/L_{Edd}$ in \citet{wan04}, although
94: true, is a trivial consequence of the fact that $Q$ is very weakly
95: correlated with $L_{bol}$ in quasars (the subpopulation of blazars
96: in their sample that excluded BL-Lacs) and not unexpectedly,
97: $L_{bol}/L_{Edd}$ and $L_{bol}$ are strongly correlated in quasars
98: \citep{tin05}. Thus, it was straightforward by means of a partial
99: correlation analysis to demonstrate that inverse correlation between
100: $Q/L_{bol}$ and $L_{bol}/L_{Edd}$ is a spurious correlation
101: \citep{tin05}.
102: \par Clearly, powerful extragalactic jets occur in high accretion thermal luminosity systems, the FRII quasars. In
103: fact, some powerful FRII quasars have $L_{bol} > 10^{47}
104: \mathrm{ergs/sec}$ \citep{tin05}. The fundamental question is
105: whether low luminosity accretion flows can also power FRII jets. The
106: answer to this question will constrain the dynamics of plausible
107: central engines that can power FRII jets. In order to shed more
108: light on these matters, we have actively searched for FRII radio
109: sources with direct, unambiguous observational evidence of a low
110: luminosity accretion flow below the Seyfert 1/QSO dividing line: a
111: total thermal luminosity (IR to X-ray) of the accretion flow,
112: $L_{bol}< 2\times 10^{45}\mathrm{ergs/s}$ (which is shown in section
113: 3.2 to be equivalent to the conventional dividing line between
114: Seyfert 1 and QSO broad-line objects, $M_{V}=-23$). To accomplish
115: this, one requires strong evidence for a direct line of sight to the
116: nucleus as in a nearly pole-on view, such as a blazar line of sight.
117: Secondly, it is preferable to have two metrics of $L_{bol}$, since
118: the optical/UV continuum is often highly contaminated by the high
119: frequency synchrotron tail of the jet emission and is at best an
120: upper bound to the optical/UV flux from the accretion flow. The line
121: strength of one or more broad UV emission lines is a more direct way
122: to measure the accretion flow luminosity in blazars. Previously, we
123: reported on the most extreme blazar in this family, 3C 216 with
124: $L_{bol}\gtrsim 10^{44}\mathrm{ergs/s}$ and a long term time
125: averaged jet kinetic luminosity, $\overline{Q}>
126: 10^{46}\mathrm{ergs/s}$, \citep{pun07}. In fact it was shown that
127: $\overline{Q}> L_{Edd}$ in 3C 216. \par This article describes the
128: broadband properties of the core dominated radio source PKSJ
129: 1037-2705. It was previously identified as a broadline AGN at
130: z=0.567 from an optical spectrum with uncalibrated flux
131: \citep{gut06}. The radio observations presented in section 2 of this
132: paper, indicate lobe emission with a 5 GHz luminosity typical of an
133: FRII radio source. The implication is that the time averaged jet
134: kinetic luminosity, $\overline{Q}\sim 5.0\times
135: 10^{44}\mathrm{ergs/s}$ (In this paper we assume: $H_{0}$=70
136: km/s/Mpc, $\Omega_{\Lambda}=0.7$ and $\Omega_{m}=0.3$). In section
137: 3, new calibrated optical observations are presented that indicate a
138: broad MgII emission line ($\sim 4000$ km/s), and the line strength
139: is consistent with a radio source that has a very low accretion flow
140: thermal luminosity, $L_{bol}\gtrsim 10^{44}\mathrm{ergs/s}$ (well
141: below the Seyfert 1/QSO dividing line). If it were not for the
142: powerful jet seen in a pole-on orientation, PKSJ 1037-2705 would be
143: a very ordinary Seyfert 1 galaxy: $L_{bol}\gtrsim 10^{44}
144: \mathrm{ergs/s}$.
145: \section{Radio Observations}The Australia Telescope Compact Array (ATCA) was used to obtain
146: observations of PKSJ 1037-2705 on November 28, 2002. The source was
147: simultaneously observed at frequencies of 4.8 and 8.64 GHz for 12
148: hours in the 6A array configuration. The bandwidth was 128 MHz at
149: both frequencies, dual polarization. The radio maps at 4.8 GHz and
150: 8.64 GHz are shown in figure 1.
151: \begin{figure*}
152: \includegraphics[width=85 mm, angle= -0]{f1.eps}
153: \includegraphics[width=85 mm, angle= -0]{f2.eps}
154: \caption{ATCA radio maps at 4800 and 8640 MHz, from left to right.
155: At 4800 MHz, the beam-size is 1.75"$\times$ 3.55" at a
156: position angle of $-1.5^{\circ}$. Contour levels for the Stokes
157: I emission are 1.26 mJy/beam $\times$ (-0.00125, 0.00125, 0.0025, 0.005, 0.01, 0.02,
158: 0.04, 0.08, 0.16, 0.32, 0.64). The peak fractional polarization is 26.3\%. The vector lengths
159: represent 5.7\% fractional polarization per arcsecond. At 8640 MHz, the beam-size is 1.03"$\times$ 1.92" at a
160: position angle of $-1.5^{\circ}$. Contour levels for the Stokes
161: I emission are 1.07 mJy/beam $\times$ (-0.005, 0.005, 0.01, 0.02,
162: 0.04, 0.08, 0.16, 0.32, 0.64). The peak fractional polarization is 23.9\%. The vector lengths
163: represent 8.2\% fractional polarization per arcsecond.}
164: \end{figure*}
165: The components were segregated as an unresolved core, an eastern jet
166: and a northern diffuse region. The core flux density distribution
167: was modeled as a single Gaussian component. The jet and the northern
168: component flux densities were integrated by hand: i.e., the residual
169: flux density after subtracting off the Gaussian core from the image
170: was summed to the east (identified as the jet) and to the north. The
171: component fluxes in the 4800 MHz image are: core flux density =
172: $136.1 \pm 13.6$ mJy, jet flux density = $25.7 \pm 2.6$ mJy and the
173: northern diffuse flux density = $8.2\pm 0.9$ mJy. At 8640 MHz, the
174: component fluxes are: core flux density = $116.3 \pm 11.6$ mJy and a
175: jet flux density = $13.8 \pm 1.4$ mJy. The error is given by the
176: $10\%$ absolute flux calibration error added in quadrature to the
177: rms noise. Modeling the jet flux density distribution as a single
178: Gaussian gave a similar result. Similarly, summing the cleaned
179: components in the u-v plane reproduced the hand integrations in the
180: northern diffuse region and the jet to within a few percent. The
181: northern diffuse emission was not detected at 8.64 GHz. The northern
182: diffuse flux density in the 8.64 GHz map is $< 1.7 $ mJy (the
183: 3$\sigma$ rms noise). Defining the radio spectral index, $\alpha$,
184: as $F_{\nu}\propto\nu^{-\alpha}$, we find that the core is flat
185: spectrum $\alpha =0.3$. The jet is extremely steep spectrum, for a
186: jet, $\alpha=1.1$, over 90\% of radio jets have $0.5 < \alpha < 0.9$
187: \citep{bri84}. Furthermore, the northern diffuse component is
188: unrealistically steep, $\alpha>2.5$. Clearly, significant amounts of
189: optically thin flux is missing due to a lack of short interferometer
190: spacings at 8.64 GHz (and possibly at 4.8 GHz as well). Therefore,
191: the jet and diffuse component spectral indices are subject to some
192: systematic uncertainties that we cannot easily quantify. The
193: complete absence of diffuse flux at 8.64 GHz to the north in a
194: relatively small (for a lobe) region, $\sim $ 5 arcsec across, makes
195: it likely that the northern component is very steep spectrum,
196: regardless of any systematic uncertainties. Strictly, speaking, the
197: eastern jet and northern diffuse component flux density should be
198: considered as lower limits, even at 4.8 GHz.
199: \subsection{A Component Model of the Radio Emission}
200: In order to assess the amount of diffuse flux that was missed by the
201: sparse array spacings, we consider the low frequency part of the
202: radio spectrum from archival TXS survey data.
203: \begin{figure*}
204: \includegraphics[width=105 mm, angle= -90]{f3.eps}
205: \caption{A component model for the broadband radio spectrum of PKSJ
206: 1037-2705. The radio data is drawn from our observations, the NASA
207: Extragalactic Database and archival literature. If the error bars
208: are smaller than the size of the marker for the observation then
209: they are not indicated in the figure. The model consists of a flat
210: spectrum radio core ($\alpha= -0.15$), a kpc scale jet ($\alpha =
211: 0.70$) and a pair of steep spectrum lobes ($\alpha = 1.25$) with
212: equal flux densities (assumed bilateral symmetry, to first order).
213: The steep lobe spectrum seems to be required by the 1.4 GHz data
214: combined with the 365 MHz data}
215: \end{figure*}
216: Figure 2 is a component model that was created to explain the
217: broadband radio data. Before going into the details of the model,
218: the salient point is that there is a huge excess of low frequency
219: emission that cannot be explained without a significant steep
220: spectrum component. The figure contains all known archival data
221: points except the PMNJ survey point which was deleted for the sake
222: of visual clarity, but is discussed below. The need to produce a
223: detailed model of the components is rather apparent from the results
224: of a direct extrapolation of the 4.8 GHz data to lower frequency.
225: Using the two point spectral indices (from 4.8 GHz to 8.64 GHz) and
226: the component fluxes, one predicts a 1.4 GHz flux density for PKSJ
227: 1037-2705 that is over 100\% larger than is observed in (see figure
228: 2). A more sophisticated analysis is in order.
229: \subsubsection{The Unresolved Radio Core} First of all, the archival data clearly
230: indicates a variable core. We note that our previous 4.8 GHz VLA,
231: C-Array (relatively low resolution) data in \citet{cla99} indicated
232: a total flux density of 241 mJy about 40\% larger than that inferred
233: from our ATCA observation ($170 \pm 17$ mJy) and the archival PMNJ
234: data ($175 \pm 14$ mJy) in NED. Such large radio variability is
235: characteristic of blazars. Since the radio data is not simultaneous,
236: any model of the core will be inaccurate subject to the scatter
237: associated with the blazar variability. Clearly, the simultaneous
238: (to the 5 GHz) 1.4 GHz C-Array VLA data form \citet{cla99} of 210
239: mJy shows that the core had an inverted spectrum (contrary to our
240: present measurement) at that time. The relatively large flux density
241: at 10.7 GHz relative to 2.7 GHz flux density found by \citet{and88},
242: considered in conjunction with the steep spectrum lobe and jet
243: components, strongly suggests that the core spectrum was inverted at
244: that epoch as well. Thus, for the model to be an "average"
245: representation over time, we chose a slightly inverted spectrum,
246: $\alpha= -0.15$ and a flux density of 136 mJy as we measured with
247: ATCA.
248:
249: \subsubsection{The Eastern Jet/Lobe Emission} Any reasonable choice of jet model will not affect our
250: conclusion that the 365 MHz data is only explained by a
251: predominantly steep spectrum component that is modest or small at
252: higher frequencies. It is curious that the jet emission is is
253: extremely steep spectrum. A jet spectral index above 1.0 is
254: extremely rare \citep{bri84}. The 4.8 GHz map in Figure 1 reveals
255: some excess diffuse emission to the north of the jet axis. Thus it
256: is likely best to think of the eastern jet as the eastern jet/lobe
257: emission. A significant amount of diffuse steep spectrum flux that
258: is co-spatially projected onto the sky plane with the jet would help
259: to explain the anomalously steep spectrum of the putative eastern
260: jet. Besides the steep spectral index, there is other evidence that
261: a portion of the eastern emission is from a lobe seen along the jet
262: axis, namely the peak of the polarization in figure 1 is not at the
263: end of the "linear" feature (the terminus of the linear feature is
264: $\approx 85-90$ kpc east of the core: there are 7.8 kpc/arsec at
265: this redshift for our adopted cosmology) as would be expected if the
266: jet terminated in a hot-spot \citep{bri94,bar90}. To the contrary,
267: there is virtually no detected polarized flux at the end of the
268: linear feature. Instead a highly polarized region is located within
269: $\approx 55$ kpc east of the core. The magnetic field is orthogonal
270: to the jet direction as is indicative of a strong knot in an FRII
271: jet or a terminating hot-spot coincident with the total intensity
272: peak, $\approx 55 - 60$ kpc to the east of the core
273: \citep{bri94,bar90}. Thus, the radio map is consistent with a hot
274: spot $\approx 55-60$ kpc from the core that terminates a pole-on
275: jet. The lobe might also be seen pole-on. In this interpretation,
276: the projection of the lobe (that is fed by the eastern jet) onto to
277: the sky plane in the 4.8 GHz map is a diffuse circular patch
278: $\approx 55$ kpc in diameter that is centered $\approx 55- 60$ kpc
279: slightly north of due east from of the core. The data is consistent
280: with this interpretation, but not definitive. In any eventuality,
281: the true jet flux is the total eastern flux minus the eastern lobe
282: flux (which we will discuss below). Approximately 40\% of kpc scale
283: jets have spectral indices between 0.6 and 0.7. Thus, for the
284: residual jet emission we picked a spectral index of 0.7 in the
285: composite model in figure 2. Again, any reasonable choice of jet
286: model will not affect our conclusion that the 365 MHz data is only
287: explained by a predominantly steep spectrum component that is modest
288: or small at higher frequencies.
289: \subsubsection{Lobe Emission}PKSJ 1037-2705 is likely a standard core
290: dominated radio source that is being viewed pole-on making the
291: de-projection of components complicated. The variable, flat-spectrum
292: core typically represents a jet beamed toward earth and the eastern
293: kpc-scale jet might also be beamed towards earth. The TXS 365 MHz
294: observation constrains the lobe flux and can be used in conjunction
295: with the high frequency radio maps to construct a model of the lobe
296: emission. The radio maps are produced from sparse array spacings so
297: we have likely missed extended flux at 4.8 GHz and almost certainly
298: at 8.64 GHz. Considering this circumstance in conjunction with our
299: inability to extricate lobe flux superimposed on the jet (or
300: co-spatial in sky plane projection with the core), the 8.2 mJy is
301: clearly a lower bound on the extended flux density at 4.8 GHz.
302: \par At a minimum, there is $8.2$ mJy of diffuse flux (to the north
303: of the core) in the 4.8 GHz map. The map seems to indicate a diffuse
304: lobe $\approx 55$ kpc in diameter. Based on the morphology of
305: blazars, it is not unusual for the lobe emission on the counter-jet
306: side to appear displaced on the sky plane roughly orthogonal to the
307: jet axis \citep{ant85}. We emphasize that there is likely more lobe
308: flux than this that is either hard to extricate from the brighter
309: features or was possibly missed due to the sparse array spacings. If
310: the source has bilateral symmetry, 16.4 mJy = $2 \times 8.2$ mJy
311: might be a more appropriate estimate for the lobe flux density (half
312: from the near lobe and half from the far lobe). The near lobe on the
313: jet side is not readily discernible, perhaps its projection onto the
314: sky plane could be confused with the powerful core or with the
315: bright eastern jet. A co-spatial projection onto the sky plane of an
316: extremely steep diffuse lobe would explain the incredibly steep jet
317: spectrum.
318: \par It is interesting to compare this conservative estimate of
319: 16.4 mJy of 4.8 GHz extended emission with the TXS data point in
320: figure 2. The absence of an extended flux detection at 8.64 GHz to
321: the north of the radio core strongly suggests that this is very
322: steep spectrum lobe emission. The sources in the 3C catalog that
323: have steep spectrum lobes, typically have $\alpha \gtrsim 1$ between
324: 750 MHz and 5 GHz \citep{kel69}. Using 3C radio sources as a guide,
325: we don't expect the spectral index from 5 GHz to 151 MHz to be any
326: steeper than 1.25 (see the 3C 368 data in NED). The same range
327: ($1<\alpha<1.25$) seems to hold with the steep spectrum 7C catalog
328: sources \citep{wil99}. Thus, we pick $\alpha = 1.25$ for the PKSJ
329: 1037-2705 lobe emission with 16.4 mJy at 4.8 GHz. After these
330: assignment of component flux densities and spectral indices, the fit
331: to the broad band data in figure 2 is reasonable considering the
332: variable core flux density and the large error in the TXS data.
333: There is no other way to explain the TXS data and the 1.4 GHz data
334: in figure 2 without a steep spectrum lobe component.
335: \subsection{Estimating The Jet Kinetic Luminosity}
336: Ostensibly, one can use the jet emission from the parsec scale radio
337: core to estimate, $Q$ at an epoch of emission that is within a few
338: years of the epoch of accretion flow emission as in \citet{cel97}.
339: Considering the potential variability in radio loud AGN (which could
340: be large if the epochs are well separated in time, especially in
341: blazars), simultaneous observations in all bands would be ideal for
342: most accurate assessment of the relation between $L_{bol}$ and $Q$.
343: Unfortunately, such estimates are prone to be very inaccurate. One
344: is observing a very small amount of dissipated energy such as X-ray
345: or optical emission as the powerful radio jet propagates away from
346: the source. One must then try to figure out the small fraction of
347: $Q$ that is dissipated in this region. Typically, the X-ray, radio
348: and optical regions are observed with different spatial resolution,
349: so it is unclear if one is detecting the same physical region on
350: parsec scales as one synthesizes the broad band data. In cases in
351: which there is sufficient broad band flux to make an estimate, one
352: is plagued with the further ambiguity of determining the Doppler
353: factor of the relativistic jet. This is a critical obstacle because
354: the luminosity from an unresolved region scales with the Doppler
355: factor to the fourth power \citep{lin85}. The situation is actually
356: worse in practice when studying blazars such as PKSJ 1037-2705 (see
357: the following sections for an expose of the blazar properties). The
358: large Doppler enhancement in blazars makes them highly variable in
359: virtually all bands. It is logistically very difficult to get
360: simultaneous broad band measurements of blazars and in fact this ia
361: rarely achieved. Thus, typically all the bands are sampled at
362: different epochs, so the error induced by the variability is imposed
363: on an already suspect method. More shortcomings of this method are
364: discussed in \citet{tin05} and a published example in which the $Q$
365: is apparently over-estimated, using radio core properties, by three
366: orders of magnitude is discussed explicitly.
367: \par The most accurate estimates of $Q$ should use an isotropic
368: estimator such as the radio lobe flux. The sophisticated calculation
369: of the jet kinetic luminosity in \citet{wil99} incorporates
370: deviations from the overly simplified minimum energy estimates into
371: a multiplicative factor $f$ that represents the small departures
372: from minimum energy, geometric effects, filling factors, protonic
373: contributions and low frequency cutoff. The quantity, $f$, is argued
374: to be constrained between 1 and 20. In \citet{blu00}, it was further
375: determined that $f$ is most likely in the range of 10 to 20. Thus,
376: choosing a value of $f=15$, \citet{pun05} converted the analysis of
377: \citet{wil99} to the formula in (2.1), even though it is just a time
378: average. This formula is an isotropic method that allows one to
379: convert 151 MHz flux densities, $F_{151}$ (measured in Jy), into
380: estimates of $\overline{Q}$ (measured in ergs/s):
381: \begin{eqnarray}
382: && \overline{Q} \approx 1.1\times
383: 10^{45}\left[(1+z)^{1+\alpha}Z^{2}F_{151}\right]^{\frac{6}{7}}\mathrm{ergs/s}\;,\\
384: && Z \equiv 3.31-(3.65) \nonumber \\
385: &&\times\left(\left[(1+z)^{4}-0.203(1+z)^{3}+0.749(1+z)^{2}
386: +0.444(1+z)+0.205\right]^{-0.125}\right)\;,
387: \end{eqnarray}
388: where $F_{151}$ is the total optically thin flux density from the
389: lobes (i.e., \textbf{no contribution from Doppler boosted jet or the
390: radio core}). The appropriate application of this equation requires
391: that one must extricate the diffuse lobe emission from the Doppler
392: boosted core and jet. The expression, (2.1), requires 151 MHz flux
393: densities, so we extrapolate the 4.8 GHz data using the same lobe
394: flux model that worked so successfully in figure 2. We
395: conservatively bound our estimate of the 151 MHz flux, by $8.2$ mJy
396: at 4.8 GHz with $\alpha^{4800}_{151} = 1.0$ at the low end and
397: $16.4$ mJy at 4.8 GHz with $\alpha^{4800}_{151} = 1.25$ at the high
398: end: $2.2\times 10^{44}\mathrm{ergs/sec}<\overline{Q}<9.3\times
399: 10^{44}\mathrm{ergs/sec}$. We remind the reader that the lower bound
400: of 8.2 mJy at 4.8 GHz is unrealistically conservative based on
401: figure 2, so this is an extremely conservative lower bound.
402: \par Alternatively, we can use the independently derived isotropic estimator
403: from \citet{pun05} which is based on the assumption that the lobe
404: material is dominated by thermal energy and the magnetic energy
405: contributions are small:
406: \begin{eqnarray}
407: &&\overline{Q}=Q_{par}\approx
408: 5.7\times10^{44}(1+z)^{1+\alpha}Z^{2}F_{151}\,\mathrm{ergs/sec}\;,\quad\alpha\approx
409: 1\;.
410: \end{eqnarray}
411: Using the same bounds on the 151 MHz flux density as above, (2.3)
412: implies, $8.8\times 10^{43}\mathrm{ergs/sec}<\overline{Q}< 4.7\times
413: 10^{44}\mathrm{ergs/sec}$. This is a conservative estimate because
414: it ignores any possible protonic component to the lobe energy.
415: Again, we note that the lower bound of 8.2 mJy at 4.8 GHz is
416: unrealistically conservative based on figure 2.
417: \par It might be a preference to use the isotropic estimators
418: directly without any reference to a particular model. Thus, we still
419: need to estimate $F_{151}$ from the 365 MHz data. Our best estimate
420: for the spectral index between 151 MHz and 365 MHz is the two-point
421: spectral index that is derived from the 365 MHz and 1.4 GHz data,
422: $\alpha=0.83$. This spectral index yields 1.38 Jy at 151 MHz which
423: translates into $\overline{Q}=8.7\times 10^{44}\mathrm{ergs/sec}$
424: and $\overline{Q}=4.3\times 10^{44}\mathrm{ergs/sec}$ from the
425: estimators in (2.1) and (2.3), respectively. The various isotropic
426: estimation techniques tend to indicate that $\overline{Q}\geq
427: 10^{44}\mathrm{ergs/sec}$. Most FRII quasars in deep surveys have,
428: $\overline{Q} \sim 10^{44}\mathrm{ergs/sec}$, thus even the most
429: conservative of isotropic estimations of $\overline{Q}$ establishes
430: an FRII level of lobe luminosity in PKSJ 1037-2705 \citep{pun01}.
431: However, the only realistic estimation models of $\overline{Q}$ are
432: those that conform with the TXS 365 MHz data. The most reasonable
433: explanation of the 664 mJy data point from the TXS survey in figure
434: 2 is that it is diffuse steep spectrum optically thin emission. One
435: could conjecture that it arose from a serendipitous observation of
436: an extremely large radio flare of the core. The required magnitude
437: of the relative increase in flux would rank this an an extreme flare
438: even by blazar standards \citep{kuh81}. If we restrict our attention
439: to the more reasonable scenarios in which the 365 MHz excess is not
440: due an extremely large flare in the core flux at low frequency then
441: the range of isotropic estimates in this section are $ 4.3 \times
442: 10^{44}\mathrm{ergs/sec}<\overline{Q}< 9.3 \times
443: 10^{44}\mathrm{ergs/sec}$.
444: \subsection{Summary of Radio Data Reduction} Our ATCA and the
445: \citet{cla99} VLA observations in conjunction with archival radio
446: data seems to indicate the following for PKSJ 1037-2705
447: \begin{itemize}
448: \item The radio source resides in a class of objects commonly
449: referred to as blazars. This deduction follows from these facts.
450: First, the radio map of PKSJ 1037-2705 is dominated by a powerful
451: unresolved radio core. Secondly, the flux density of the core shows
452: at least 40\% variability on the time scale of a few years. Thirdly,
453: the spectral index is the core is flat spectrum ($\alpha<0.5$) and
454: it is variable changing sign over time. All of these facts are
455: consistent with a relativistic jet beamed towards earth.
456: \item There is an excess of flux at 365 MHz that can not be
457: explained by the core and jet flux alone.
458: \item There is resolved steep spectrum lobe emission in the 4.8 GHz
459: map
460: \item The 1.4 GHz and 365 MHz data imply that the 365 MHz flux
461: density is comprised mainly of $>400$ mJy of very steep spectrum
462: emission that is most likely associated with the lobe flux.
463: \item The large amount of lobe flux at low frequency requires a
464: large amount of stored magnetic plasma energy within the lobes.
465: Independent isotropic estimators of this energy density indicate
466: that the most reasonable explanation of the lobe energy density
467: requires a jet with a time averaged kinetic luminosity, $ 4.3 \times
468: 10^{44}\mathrm{ergs/sec}<\overline{Q}< 9.3 \times
469: 10^{44}\mathrm{ergs/sec}$. - an FR II level of power.
470: \end{itemize}
471: \section{The Optical Observations}PKSJ 1037-2705 was observed with 3x1200 sec exposures using the
472: Telescopio Nazionale Galileo (TNG) telescope with the LR-B grism on
473: February 24, 2007. The spectral reduction, wavelength, flux
474: calibration were achieved using standard IRAF procedures (see Figure
475: 3)\footnote{IRAF is the Image Reduction and Analysis Facility,
476: written and supported by the IRAF programming group at the national
477: Optical Astronomy Observatories (NOAO) in Tucson, Arizona.}. The
478: slit width was 1.5". The spectrum was smoothed with a squared box of
479: 5 pixels. The calibration in flux was done relative to the star BD75
480: +325 (spectral type O5). Wavelength calibration was done with a He
481: lamp. The spectrum in Figure 3 is corrected for Galactic extinction
482: using the values given in \citet{sch98}. The most striking feature
483: is that the the continuum is very steep, $\alpha = 3.16$ (from 7500
484: \AA to 4500 \AA). The H$\beta$ emission is blocked by telluric band
485: absorption. The H$\gamma$ line is positioned just within the blue
486: side of the telluric band at 6800 \AA. However, we do clearly detect
487: one broad emission line (BEL), at $4399.35\AA$ that we have
488: identified as MgII. Our IRAF Gaussian fit yields an intrinsic MgII
489: line strength, $L_{MgII} \approx 1.2 \times
490: 10^{42}\mathrm{ergs/sec}$ and the FWHM is $\approx 4000 $ km/s. The
491: procedure to get these values is described below.
492:
493: \begin{figure}
494: \includegraphics[width=150 mm, angle= 0]{f4.eps}
495: \caption{The optical spectrum of PKSJ 1037-2705 observed with TNG
496: corrected for Galactic extinction. All extinction values in this
497: paper are from \citet{sch98}.}
498: \end{figure}
499: \begin{figure}
500: \includegraphics[width=150 mm, angle= 0]{f5.eps}
501: \caption{A closeup of the TNG spectrum of PKSJ 1037-2705 in figure 3
502: in the vicinity of Mg II. The horizontal red lines indicate the one
503: sigma uncertainty in the continuum level near the Mg II BEL ($ 4200
504: \AA < \lambda < 4600 \AA$) caused by noise. We present a range of
505: Gaussian fits to the Mg II BEL as the continuum level is varied from
506: the average in the band to $\pm 1 \sigma$. The thick curve is the
507: fit to the average continuum and the thin curves are the fits to the
508: average continuum $\pm 1 \sigma$. The bottom frame shows the
509: residuals to our Gaussian fits, it looks similar to the noise in the
510: rest of the band.}
511: \end{figure}
512: \subsection{The MgII Broad Emission Line} The properties of the MgII
513: line are important to the discussion to follow. Unfortunately, the
514: continuum above and below the BEL is not well constrained due to
515: significant noise and there is apparently significant noise that
516: distorts the line shape near the peak. Since this measurement is
517: important in the following analysis, we take special care to
518: quantify the uncertainties in the measurement of the line strength
519: and FWHM.
520: \subsubsection{The Continuum Level} The major source of error in our estimates
521: arises from uncertainties in the continuum level. Since we don't
522: expect any other emission lines nearby, the first thing that we can
523: look at is the average continuum level both red-ward and blue-ward
524: of the BEL. The errors that are induced by the noise were obtained
525: by sampling the blue side of the Mg II BEL from 4200 $\AA$ to 4300
526: $\AA$ and on the red side from 4500 $\AA$ to 4600 $\AA$. We combined
527: the red and the blue bands to estimate the local continuum level as
528: $3.13 \pm 0.21 \times 10^{-17} \mathrm{ergs/s/cm^{2}/\AA}$. This
529: band of one sigma uncertainty in the local continuum level is
530: superimposed on the spectrum in figure 4, by two red horizontal line
531: segments.
532: \subsubsection{The Gaussian Fits} Figure 4 shows the single Gaussian fits
533: to the data with the continuum set at the three levels noted in the
534: subsection above, nominal (the average continuum level) and nominal
535: $\pm 1 \sigma$. A single Gaussian fit to the line with the continuum
536: level fixed is given by option "h" in IRAF splot. The IRAF generated
537: fits have residuals that look consistent with the noise level
538: surrounding the BEL. The FWHM values were extracted from the fitted
539: Gaussian models after correcting for instrumental broadening which
540: is about 20 $\AA$ measured from the sky and arclines. Our analysis
541: indicates that the three single Gaussian fits give a range of
542: uncertainty from the single Gaussian fit to the nominal continuum
543: that is expressed as $L_{MgII} = 1.23^{+.37}_{-.41} \times
544: 10^{42}\mathrm{ergs/sec}$ and $\mathrm{FWHM}=4266^{+321}_{-570}$
545: km/s. If we consider a conservative 20\% uncertainty in absolute
546: flux calibration (this affects $L_{MgII}$, but not to the FWHM), we
547: compute that $L_{MgII} = 1.23^{+.69}_{-.57} \times
548: 10^{42}\mathrm{ergs/sec}$.
549:
550: \subsubsection{The Non-Gaussian Fit} A single Gaussian fit might not be
551: the best model of the MgII BEL in PKSJ 1037-2705. There is a
552: significant red-ward asymmetry and there is an irregular peak
553: possibly (the significant excess residual in figure 4) from a strong
554: narrow line component (see section 3.5 for the evidence of a strong
555: narrow line emissivity that is excited by the propagating jet).
556: Another possible reason for the distortion from a smooth profile
557: near the line peak is that the MgII line is actually a doublet.
558: Thus, we are interested in estimating $L_{MgII}$ and the FWHM
559: without assuming any particular parametric form (such as a single
560: Gaussian). The IRAF option "e" allows us to directly integrate the
561: flux above a certain continuum level and this seems suitable for our
562: purposes. Given the irregular line shape this direct approach is
563: preferable to the single Gaussian model for computing $L_{MgII}$.
564: Again, we adopt a conservative 20\% uncertainty in absolute flux
565: calibration and our direct integration for each of the three levels
566: of continua in Figure 4 yield a spread of values about the nominal
567: continuum estimate of $L_{MgII} = 1.24^{+.58}_{-.48} \times
568: 10^{42}\mathrm{ergs/sec}$. Which is very close to the value and
569: uncertainty found from the single Gaussian fits in the previous
570: subsection.
571: \par We compute the FWHM in a similar manner with the direct
572: integration in IRAF option "e". We directly measured the width (in
573: $\AA$) of the spread in the data halfway between the peak and the
574: continuum. We corrected for instrumental broadening as in the
575: previous subsection and found a spread in the FWHM estimates
576: relative to the value found using the nominal continuum given by
577: $\mathrm{FWHM}=3504^{+313}_{-799}$ km/s. The nonparametric fit and
578: the Gaussian fits give values of the FWHM that are different at the
579: 1.2 $\sigma$ level.
580:
581: \subsection{Estimating $L_{bol}$ from the Continuum
582: Spectrum} The total bolometric luminosity of the accretion flow,
583: $L_{bol}$, is the thermal emission from the accretion flow,
584: including any radiation in broad emission lines from photo-ionized
585: gas, from IR to X-ray. Since we do not have complete broadband
586: coverage of the SED for the accretion flow, one can estimate
587: $L_{bol}$ by means of a composite SED. The most important samples
588: for the composite quasar spectra are the HST observations since
589: these cover the rest frame EUV (extreme ultraviolet) region in which it was
590: once believed that much of the quasar energy was hidden
591: \citep{zhe97}. Figure 5 is a composite spectral energy distribution
592: of a radio quiet quasar with $M_{V}=-25$ (roughly the average value
593: in the \cite{zhe97} HST sample). This spectrum, in combination with
594: the broad emission lines, represents the ``typical'' radiative
595: signature of a strong accretion flow onto a black hole in the
596: absence of an FR II jet. This signature is empirical and is
597: independent of all theoretical models of the accretion flow.
598: \begin{figure}
599: \plotone{f6.eps}
600: \caption{The composite spectral energy distribution of a quasar
601: with $M_{V}=-25$ from the combined data of
602: \citet{elv94,zhe97,tel02,lao97}}
603: \end{figure}
604: \par Figure 5 is a piecewise collection of power laws that approximate
605: the individual bands. The IR and optical data are from the
606: composite spectrum of \citet{elv94}, the NUV (near ultraviolet)
607: and EUV data are from the HST composites of \citet{zhe97,tel02}
608: and the X-ray portion of the composite is from \citet{lao97}. In order to compute
609: the bolometric luminosity of the
610: accretion flow from the composite spectral energy
611: distribution, one can estimate (from the tables in \citet{zhe97}and \citet{tel02}) that 25 percent of
612: the total optical/UV quasar luminosity is reprocessed in the broad
613: line region. Thus, for this $M_{V}=-25$ radio quiet composite SED,
614: $L_{bol}=1.35\times 10^{46}\mathrm{ergs/sec}$. If one also
615: assumes that the shape of the composite spectrum in Figure 5 is
616: independent of quasar luminosity, a simple approximate formula
617: is obtained that relates the k-corrected absolute visual magnitude with bolometric
618: luminosity,
619: \begin{eqnarray}
620: && L_{bol}\approx 1.35 \times 10^{\frac{-25-M_{V}}{2.5}}\times
621: 10^{46}\mathrm{ergs/sec}\;.
622: \end{eqnarray}
623: As a more general alternative to (3.1), if $L(\nu)_{\mathrm{obs}}$
624: is the observed spectral luminosity at the AGN rest frame frequency,
625: $\nu$, then $L_{bol}$ is estimated as
626: \begin{eqnarray}
627: && L_{bol}=1.35\frac{\nu L(\nu)_{\mathrm{obs}}}{\nu
628: L(\nu)_{\mathrm{com}}}\times 10^{46}\mathrm{ergs/sec}\;,
629: \end{eqnarray}
630: where $L(\nu)_{\mathrm{com}}$ is the spectral luminosity from the
631: composite SED.
632: \par The problem with applying (3.1) and (3.2) to
633: PKSJ 1037-2705 is that one must subtract off the high frequency tail
634: of the synchrotron jet. A strong contribution from the jet that
635: masks the optical signature of the accretion flow seems likely given
636: the best fit continuum spectral index (defined in section 2 as
637: $F_{\nu}\propto\nu^{-\alpha}$) of $\alpha^{7500}_{4500}=3.16$ (from
638: 7500 $\AA$ to 4500 $\AA$). In order to ascertain if the high
639: frequency tail of this blazar component masks the optical/near UV
640: flux from the accretion flow, we created a broadband SED (i.e., a
641: plot of $\nu L_{\nu}$ versus $\nu$) in figure 6.
642: \begin{figure*}
643: \includegraphics[width=95 mm, angle= -90]{f7.eps}
644: \caption{A broadband spectral energy distribution or SED (i.e., a
645: plot of $\nu L_{\nu}$ versus $\nu$) for PKSJ 1037-2705. The SED is
646: corrected for Galactic Extinction. The radio data is drawn from
647: figure 2. The IR data is from the 2MASS survey and was accessed
648: through MAST. The optical data is from our observations. The UV data
649: point was from the GALEX shallow sky survey and was retrieved
650: through MAST. The X-ray data points come from the fit to our
651: Newton-XMM observation shown in Figure 8 (see section 4). If the
652: error bars are smaller than the size of the data point marker then
653: they are not shown.}
654: \end{figure*}
655: Even though the data in the SED in Figure 6 were not obtained
656: simultaneously (and blazars are highly variable), the basic shape is
657: quite revealing. First of all, notice the peak in the SED in the IR.
658: The peak of the SED is clearly in excess of
659: $10^{45}\mathrm{ergs/sec}$ in the Mid-IR. The spectrum from
660: $\lesssim 10^{10}$ Hz to $\gtrsim 10^{15}$ Hz is dominated by one
661: component with a shape typical of the synchrotron spectrum of a
662: blazar jet \citep{fos97}. Compare this spectrum to the signature of
663: the accretion disk, the SED for radio quiet quasars in Figure 5. The
664: thermal SED from the accretion flow is rising from $3\times 10^{14}$
665: Hz to $3\times 10^{15}$, the "big blue bump". Conversely, the SED
666: for PKSJ 1037-2705 is a steeply decreasing function in this same
667: frequency range, a clear distinction from a thermal origin. The rise
668: of the SED toward high frequency in the X-ray regime is most likely
669: the inverse Compton emission from this very same jet (see the next
670: section for details) which is another characteristics of blazar jets
671: \citep{fos97}. There is no evidence of a break in the synchrotron
672: spectral index (the optically thin portion of the synchrotron peak)
673: from the IR to the near UV (see section 3.3). The GALEX data has a
674: large error bar associated with it, so it is not too useful. All
675: that the GALEX data really tells us is that there is no strong
676: strong thermal component in the UV.
677: \par The SEDs in Figures 5 and 6 indicate that the
678: optical component from the jet is clearly much larger than that of
679: the accretion disk. The data (the steep power law) indicates that
680: down to at least $10^{15}$ Hz, the underlying continuum is still
681: mostly hidden by the jet flux (see also Figure 7 and the discussion
682: in section 3.3). Therefore, we can use Figures 5 and 6 with equation
683: (3.2) to create an upper bound on the accretion disk luminosity. Due
684: to the rapidly falling SED, the most stringent upper bound is formed
685: by choosing the highest frequency UV point with relatively small
686: errors associated with it (not the GALEX data), $\approx 10^{15}$
687: Hz,
688: \begin{eqnarray}
689: && L_{bol}=1.35\frac{\nu L(\nu)_{\mathrm{thermal}}}{\nu
690: L(\nu)_{\mathrm{com}}}\times 10^{46}\mathrm{ergs/sec}\nonumber\\
691: && \ll 1.35\frac{\nu L(\nu)_{\mathrm{synchrotron}} + \nu
692: L(\nu)_{\mathrm{thermal}}}{\nu L(\nu)_{\mathrm{com}}} \times
693: 10^{46}\mathrm{ergs/sec} = 7.64 \times 10^{44} \mathrm{erg/s}\;,
694: \end{eqnarray}
695: where $\nu L(\nu)_{\mathrm{thermal}}$ is the contribution to the SED
696: of PKSJ 1037-2705 from the thermal accretion flow and $\nu
697: L(\nu)_{\mathrm{synchrotron}}$ is the synchrotron contribution to
698: the SED (evaluated at $\approx 10^{15}$ Hz).
699: \par This upper bound
700: is quite conservative and it is based on not finding any true signal
701: of the accretion flow. We want to do better than this, so we dig
702: deeper into the optical data set to find more evidence of the
703: thermal accretion flow in the next subsections.
704: \begin{figure}
705: \includegraphics[width=100 mm, angle= -90]{f8.eps}
706: \caption{The residual flux density of the UV (in the quasar rest
707: frame) continuum relative to the global synchrotron power law. A gap
708: is left that accommodates the Mg II BEL that significantly distorts
709: the continuum level. The best fit, steep power law, from 7500 $\AA$
710: down to 5000 $\AA$ that is given by $\alpha = 3.39$, continues to
711: follow the continuum very closely down to 4500 $\AA$. The figure
712: indicates a hint of an excess of flux below 4300 $\AA$, but the
713: results is not statistically significant. A slight excess would be
714: indicative of a weak thermal component near 4000 $\AA$. The flux
715: density is corrected for Galactic extinction and the units are
716: $\mathrm{ergs/s/cm^{2}/\AA}$}
717: \end{figure}
718: \subsection{Estimating $L_{bol}$ Based on the UV Excess from the Global Power Law}
719: In order to make a better estimate than the loose upper
720: bound in (3.3), one can look for some weak signal that is indicative
721: of the thermal accretion flow. If there is no spectral ageing
722: effects, one expects the optically thin portion of the jet
723: synchrotron spectrum to be well fit by a power law. Since the
724: optically thin synchrotron spectrum is so steep, the best chance for
725: detecting evidence of a weak $L_{bol}$ is by looking for an
726: inflection point in the power law spectrum blue-ward of Mg II. The
727: farther into the UV we go, the more pronounced the thermal component
728: becomes, based on a typical quasar spectrum from \citet{zhe97} and
729: the synchrotron spectral index,
730: $F_{\lambda}(\mathrm{thermal})/F_{\lambda}(\mathrm{synchrotron})
731: \sim \lambda^{-2.5}$.
732: \par We note that the spectral index flattens as the end point of
733: the fit to the continuum on Figure 3 moves deeper into the UV. The
734: values of spectral index of the best fit, de-reddened spectrum for
735: an endpoint that goes from 5500 $\AA$ to 5000 $\AA$ to 4500 $\AA$ to
736: 3900 $\AA$ are: $\alpha^{7500}_{5500}=3.44$,
737: $\alpha^{7500}_{5000}=3.39$, $\alpha^{7500}_{4500}=3.16$,
738: $\alpha^{7500}_{3900}= 2.92$. It appears that there is a break in
739: the synchrotron power law short-ward of 5000 $\AA$. This could be a
740: sign of a UV excess above the synchrotron power law. This is not
741: expected for a pure synchrotron source since spectral ageing, if
742: present, actually steepens the spectrum at high frequencies. A UV
743: excess over a pure power law fit, could be a signal of an accretion
744: disk starting to become noticeable in the strong glare of the
745: synchrotron jet. Unfortunately, the spectrum is noisy in the UV, as
746: noted in the last subsection, so the excess is not statistically
747: significant.
748: \par We explore this trend toward a UV excess over a power law by
749: first fitting the continuum spectrum from 7500 $\AA$ to 5000 $\AA$,
750: $\alpha^{7500}_{5000}=3.39$. We then ask the question, if this power
751: law were extended toward the far UV, what would the residuals of the
752: data look like, i.e., the excess. The results are plotted in Figure
753: 7. There is evidence of a gradual trend for an excess extending into
754: the band from 4300 $\AA$ to 3900 $\AA$. We note that in the band
755: 4900 $\AA$ to 4500 $\AA$ that the average residual flux density is
756: $4.5 \pm 5.1 \times 10^{-18} \mathrm{ergs/s/cm^{2}/\AA}$ and in the
757: shorter wavelength band, 4300 $\AA$ to 3900 $\AA$ the average excess
758: seems to increase to $9.1 \pm 7.0 \times 10^{-18}
759: \mathrm{ergs/s/cm^{2}/\AA}$. But the excess is $\sim 1 \sigma$ above
760: the synchrotron power law, and therefore not statistically
761: significant.
762: \par We can use this lack of statistical significance
763: to obtain another upper bound on $L_{bol}$. If we use the flux
764: excess in the band 4900 $\AA$ to 4500 $\AA$ to represent the value
765: at the center of the narrow band, 4700 $\AA$ (approximately 3000
766: $\AA$ in the quasar rest frame), then we know that the
767: $F_{\lambda}(\mathrm{thermal})(4700)< 1.98 \times 10^{-17}
768: \mathrm{ergs/s/cm^{2}/\AA}$ at the 3 $\sigma$ level. Then (3.2)
769: implies that $L_{bol}< 5.5 \times 10^{44}$ ergs/s at the 3 $\sigma$
770: level. This is a tighter upper bound than was found in (3.3).
771: Similarly, one can estimate the excess from the global power law at
772: 4100 $\AA$ (2615 $\AA$ rest frame) by averaging the excess within
773: the band from 4300 $\AA$ to 3900 $\AA$. The flux in this band,
774: quoted above, inserted into equation (3.2) yields $L_{bol} =2.1 \pm
775: 1.6 \times 10^{44}$ ergs/s, which is consistent with the upper bound
776: from the 4700 $\AA$ excess.
777: \subsection{Estimating $L_{bol}$ from
778: Broad Line Strengths} In order to constrain the accretion disk
779: thermal luminosity, a different estimator is required. The estimator
780: in (3.2) of bolometric luminosity of the accretion flow is a direct
781: measurement of an accretion flow property. Thus, it is superior to
782: using line emission to estimate the accretion flow power which is an
783: indirect estimator. The advantage of using the line luminosity is
784: that its strength might not be affected by the relativistic jet to
785: first order. However, the down side of using line luminosity is that
786: it is a second order indicator of accretion luminosity. One can use
787: either broad lines which are $\sim$ 0.01 pc from the accretion disk
788: or narrow lines which can occur anywhere from just outside the broad
789: line region out to distances of $\sim 50$ kpc \citep{bes01}. First
790: of all, the broad lines are closer to the photo-ionization source,
791: so it seems that this might be a better choice. Secondly, it has
792: been shown the the narrow lines can be created by jet propagation
793: and might not provide a reliable diagnostic of the thermal
794: photo-ionization source in many radio sources \citep{bes00}. This
795: will discussed in more detail in the next subsection. That is why
796: broad lines are the first choice for blazar accretion flow
797: estimates. The method that is most commonly used for estimating
798: $L_{bol}$ in blazars is to compare the line strengths to a composite
799: SED \citep{cel97,wan04}. Again, we note that the existence of a
800: radio jet tends to produce strong narrow lines in NLRGs
801: \citep{bes00}. Since the MgII line strength is so weak, subtraction
802: of the narrow line component associated with the propagating jet is
803: likely important to accurately estimate the broad line strength,
804: $L_{MgII}(BEL)$, arising from the photo-ionization of BEL clouds by
805: the accretion flow thermal emission. The composite NLRG ($z\sim 1$)
806: spectrum produced in \citet{bes01}, indicates that the MgII narrow
807: line strength, $L_{MgII}(NL)$ is $\approx 0.15$ of the OII line
808: strength. Our IRAF single Gaussian (option k) fit to the spectral
809: data indicate that $L_{OII}(NL)\approx 2.4\times
810: 10^{42}\mathrm{ergs/sec}$. Thus, from the data reduction in section
811: 3.1.3, $ 4.0\times 10^{41}\mathrm{ergs/sec}< L_{MgII}(BEL)\approx
812: L_{MgII} - 0.15L_{OII}(NL) < 1.5\times 10^{42}\mathrm{ergs/sec}$. We
813: then use the composite SED in Figure 5 combined with the HST
814: composite quasar line strengths of \citet{zhe97} (from which the
815: composite SED is derived) to compute $L_{bol} \approx 204
816: L_{MgII}(BEL)$, which is about 20\% more than the \citet{wan04}
817: formula ($L_{bol} \approx 168 L_{MgII}(BEL)$). For PKSJ 1037-2705
818: this relation implies $ 8.2\times 10^{43}\mathrm{ergs/sec}< L_{bol}
819: < 3.1\times 10^{44}\mathrm{ergs/sec}$. This is consistent with the
820: upper bound estimated from the optical continuum, $L_{bol}< 5.5
821: \times 10^{44}$ ergs/s, that was found in the previous subsection.
822: \subsection{Estimating $L_{bol}$ from
823: Narrow Line Strengths} A narrow line estimator for $L_{bol}$ could
824: also be tried. A seminal effort in \citet{raw91} estimated that the
825: total narrow line luminosity was
826: \begin{eqnarray}
827: L_{NLR} \approx 9 L_{OII3727} + 4.5 L_{OIII5007}\;.
828: \end{eqnarray}
829: For PKSJ 1037-2705 the lines strengths from the IRAF Gaussian fits
830: to the spectrum in figure 3 are $L_{OII3727}\approx 2.4 \times
831: 10^{42}\mathrm{ergs/sec}$ and $L_{OIII5007}\approx 4.8\times
832: 10^{42}\mathrm{ergs/sec}$. Thus, from equation (3.4), we estimate as
833: per \citet{raw91}, a narrow line luminosity of $L_{NR}\approx
834: 4.3\times 10^{43}\mathrm{ergs/sec}$. In \citet{raw91} they picked an
835: arbitrary covering factor of 0.01 for the narrow line clouds (which
836: is likely too large by a factor of $\gtrsim 3$, see below and
837: \citet{wil99}). Therefore they estimated the blue bump optical/UV
838: luminosity, $L_{BB} = 100 L_{NLR}$. From the SED in Figure 5 and the
839: broad line strengths in \citet{zhe97}, $L_{bol}\approx 1.67
840: L_{BB}\approx 7.2\times 10^{45}\mathrm{ergs/sec}$.
841: \par Alternatively, \citet{wil99} claim that $L_{OIII}$ is not
842: reliable for the estimation of $L_{bol}$ and claim that their new
843: OII estimator is superior to their earlier efforts,
844: \begin{eqnarray}
845: L_{BB} \approx 5\times 10^{3} L_{OII3727}\;.
846: \end{eqnarray}
847: The \citet{wil99} estimate gives $L_{bol}\approx 1.67 L_{BB}\approx
848: 1.9\times 10^{46}\mathrm{ergs/sec}$. The primary difference between
849: the \citet{wil99} and \citet{raw91} is the adhoc covering factor.
850: The revised \citet{wil99} value is 0.003. Note that the revised
851: covering factor is much more compliant with the composite quasar
852: spectrum of \citet{fra91} that is discussed below. With the revised
853: covering factor the narrow line luminosity implies that the flux
854: density at 3000 $\AA$ from the "big blue bump," (the ionizing
855: continuum) should be 30 times larger than what we observed in the
856: spectrum of PKS J1037-2705. This begs the question, why is the
857: narrow line estimator so poor for this source? Recall that
858: \citet{bes00} showed that $L_{NLR}$ can be from gas that was excited
859: by shocks that are driven by the jet in radio galaxies. By using
860: long slit spectroscopy, they found many regions of OII emitting gas
861: that are aligned with features in the jet. A study of narrow line
862: ratios in these aligned features indicated that the excitation state
863: and density was explained best by the shock models. In fact, they
864: showed that the smaller radio galaxies ($<120$ kpc) and those with
865: broad OII lines ($\sim $ 1000 km/sec) had the majority of their
866: narrow line gas in an ionization state that was likely excited by
867: jet induced shocks as opposed to photo-ionization. Thus, $L_{NLR}$
868: is a poor diagnostic of the thermal accretion luminosity in these
869: sources. PKSJ 1037-2705, is this type of source, a linear size
870: $<120$ kpc and a single Gaussian fit to the OII line yields a FWHM
871: of $\approx $ 1000 km/sec. The narrow lines are driven primarily by
872: the jet since the photo-ionization source associated with the
873: accretion disk is so weak as evidenced by the weak MgII BEL and the
874: low level UV continuum emission. In order to quantify how weak the
875: continuum ionization source is compared to the jet shock ionization
876: source, we note that in the composite quasar spectrum of
877: \citet{fra91}, (The \citet{zhe97} quasar composite does not cover
878: OII and OIII): $ L_{OIII}/L_{MgII} =0.1$ and $L_{OII}/L_{MgII}
879: =0.023$. By contrast, for PKSJ 1037-2705, we have $L_{OIII}/L_{MgII}
880: \approx 3.9$ and $ L_{OII}/L_{MgII} \approx 1.9$ (based on our best
881: fit Gaussians). It is concluded that $L_{NLR}$ is a poor predictor
882: of $L_{bol}$ because the ratio of the jet kinetic luminosity to the
883: photo-ionizing luminosity is extremely high in this quasar. In fact,
884: it is proposed that a potential method for selecting kinetically
885: dominated quasars is to find blazars (unobscured nuclei) in which $
886: L_{OII}/L_{MgII}> 1 $ and the MgII FWHM $>$ 2000 km/sec.
887:
888: \subsection{Summary of IR/Optical Observations}In this section we
889: used optical observations with TNG and 2MASS archival IR data to
890: quantify numerous characteristics of the AGN, PKSJ 1037-2705.
891: \begin{itemize}
892: \item PKSJ 1037-2705 is blazar with a synchrotron peak luminosity
893: in the IR with $\nu L_{\nu}>10^{45}$ ergs/sec (see section 3.2).
894: \item The high frequency synchrotron tail from the blazar dominates
895: the optical and near UV luminosity created by the accretion flow
896: (see section 3.3)
897: \item PKSJ 1037-2705 has a Seyfert 1 thermal spectrum as evidenced
898: by the MgII broad emission line (see section 3.1).
899: \item Standard estimation techniques based on composite quasar
900: spectra and the broad line strength indicate that $ 8.2\times
901: 10^{43}\mathrm{ergs/sec}< L_{bol} < 3.1\times
902: 10^{44}\mathrm{ergs/sec}$. This is an order of magnitude below the
903: Seyfert 1/QSO dividing line.
904: \end{itemize}
905: \begin{figure*}
906: \includegraphics[width=125 mm, angle= -0]{f9.ps}
907: \caption{The X-ray spectrum from XMM-Newton. The different colors
908: represent data for the 3 XMM-Newton cameras: black = PN, red: MOS1,
909: blue: MOS2}
910: \end{figure*}
911: \section{The X-ray Observations}PKSJ1037-2705 was observed on July 3, 2004 for 18 ks with
912: XMM-Newton. The XMM-Newton data are affected by solar flares,
913: yielding a cleaned exposure time of only 5 ks. Following standard
914: light-curve screening of the XMM-Newton data, spectra and response
915: products were extracted for each EPIC camera within a 30 arcsec
916: radius from the X-ray peak, using a surrounding 1-2.5 arcmin
917: point-source excised annulus for background estimates. The three
918: spectra were each accumulated in bins containing at least 25 net
919: counts and then jointly fitted in XSPEC v11.3 using standard
920: $\chi^2$ minimization. Various models were fitted with Galactic
921: absorption (Comptonized black-body, Bremsstrahlung model, accretion
922: disk with multiple black-body components, thermal plasma at z=0.57
923: and a power-law). A simple power-law absorbed by the Galaxy (ie.
924: with no evidence of intrinsic absorption at the source) provides the
925: best fit among the models that we tried. The spectrum is shown in
926: Figure 8. The spectral fitting was done in the 0.4-5 keV band. We
927: find the spectral index, $\Gamma = 1.77 \pm 0.20$ with reduced
928: $\chi^{2}$=1.44 for 21 degrees of freedom (the null hypothesis
929: probability is 0.0086). Including intrinsic absorption in the model,
930: the best-fit hydrogen column density of such a component is only
931: $\sim 20\%$ of the Galactic value and consistent with zero at 1
932: sigma (and with no improvement in fit quality). As a check of the
933: veracity of the fit, fitting was also performed using Cash
934: statistics on spectra binned into 5 cts/bin, but the results were
935: generally consistent with those of $\chi^2$ fitting reported here.
936: \par The absorption-corrected X-ray luminosity in the quasar rest frame (the XMM-Newton band covered in
937: Figure 8 transforms to 0.63-7.85 keV), is $2.49^{+0.42}_{-0.35}
938: \times 10^{44}$ ergs/s. This value actually is comparable to or
939: larger than the estimates of $L_{bol}$ that was computed from the
940: broadline strength, a common signature of a blazar. Archival ROSAT
941: data of this source (also known as IXO 37) was extrapolated with
942: $\Gamma = 1.7$ (similar to our value) in \citet{col02} to compute
943: the 2 -10 keV intrinsic luminosity. After correcting for the
944: incorrect redshift in \citet{col02} and extrapolating our best-fit
945: model to to 2 -10 keV, we find that the source was $\approx 2.5$
946: times more luminous at the time of the ROSAT observations. This type
947: of large X-ray variation (a factor of a few, in time frames of years
948: or shorter) is characteristic of a blazar.
949: \section{Discussion} In this article, we established the following
950: \begin{enumerate}
951: \item PKSJ 1037-2705 is a blazar based on radio core dominance, a
952: flat radio core spectrum, the large radio and X-ray variability, a
953: steep optical spectrum and an inordinately large X-ray luminosity
954: (relative to the UV and $L_{MgII}$).
955: \item PKSJ 1037-2705 has the extended radio flux typical of an FR II
956: quasar.
957: \item PKSJ 1037-2705 is a broadline object that is likely to have an accretion
958: flow luminosity an order of magnitude weaker than a quasar, i.e. it
959: is a Seyfert 1 galaxy.
960: \item PKSJ 1037-2705 is likely to be kinetically dominated based on
961: the spread in the estimated values of $\overline{Q}$ and $L_{bol}$
962: that were determined in sections 2 and 3, respectively: $4.3/3.1
963: =1.4 < \overline{Q}/L_{bol}< 9.3/0.82 = 11.3 $.
964: \end{enumerate}
965: \par The proof that Seyfert 1 AGN can host FRII jets is based on a few
966: anecdotal examples. Note that these objects are distinct from broad
967: line radio galaxies (BLRGs). The Mid-IR observations of
968: \citet{ogl06} indicate that FRII BLRGs have nuclei with quasar level
969: luminosity which are often obscured by dusty gas. Even though
970: Seyfert 1 FRII radio sources might not be exceptionally rare,
971: demonstrating the existence of a single object explicitly is
972: extremely challenging. For example, \citet{ogl06} suggest that their
973: weak Mid-IR subsample of NLRGs are hosted by low luminosity AGN,
974: i.e., Seyfert 1 level thermal luminosity. However, \citet{cle07}
975: would argue that the absorbing columns to these NLRGs is so dense
976: that even the IR can not escape and there is still a hidden quasar
977: buried inside.
978: \par The existence of Seyfert 1 nuclei in FRII sources is important
979: for understanding the connection between the accretion flow and the
980: jet power, with two possible implications. Firstly, the kpc scale
981: emission that was used to compute $\overline{Q}$ was ejected from
982: the AGN $\sim 10^{6} - 10^{7}$ years earlier than $L_{bol}$ was
983: created \citep{wil99,blu00}. It is not clear if there should be a
984: strong connection between the present value of $L_{bol}$ and the
985: value of $L_{bol}$ $\sim 10^{6} - 10^{7}$ years ago. On the one
986: hand, strong variability has been detected in the thermal emission
987: from Seyfert 1 nuclei with changes of a factor of 5 in luminosity in
988: weeks or months \citep{ant83,all85}. The observation of this type of
989: variability has led researchers such as \citet{all85,cut85,ant88} to
990: conclude that the variability time scale for $L_{bol}$ can not be
991: related to the thermal and viscous time scales of an accretion disk.
992: On the other hand, there is evidence that some high redshift quasars
993: have been active for $> 10^{7}$ years based on the $\mathrm{He}^{+}$
994: Lyman $\alpha$ "proximity effect" \citep{jak03}. The ``proximity
995: effect'' arises when a large bubble forms about the quasar in which
996: helium is highly ionized by the quasar. One can use the size of the
997: bubble to estimate the amount of time it would take for the
998: photo-ionizing source to ionize the extended region. However, one
999: must be cautious to extrapolate this result to much less luminous
1000: Seyfert 1 galaxies and it does not provide evidence that the
1001: photo-ionization source was steady within a factor of 10 during the
1002: entire $10^{7}$ years. With all this being said, it would be a huge
1003: stretch in reasoning to assume that in spite of the large short term
1004: variations that have been observed in some Seyfert 1 galaxies that
1005: $L_{bol}$ is constant to within a factor of 10 over $10^{7}$ years.
1006: Hence, there does not seem to be any compelling observation or
1007: underlying reason why the two quantities, $\overline{Q}$ and
1008: $L_{bol}$ should be related, even if the accretion flow drives the
1009: jet.
1010: \par Alternatively, the existence of Seyfert 1 nuclei in FR II radio sources could mean that the
1011: accretion rate is not that strongly coupled to the jet power.
1012: Unfortunately, as a consequence of the Doppler boosting there is no
1013: reliable contemporaneous estimate of $Q(t)$ that can be used to test
1014: the latter hypothesis (see the discussion in section 2.2). In the
1015: context of our previous work, we have now reported on 3 Seyfert 1
1016: galaxies that have $\overline{Q} \geq L_{bol}$, PKSJ 1037-2705
1017: ($L_{bol}\gtrsim 10^{44}$ ergs/s), 3C 216 ($L_{bol}\gtrsim 10^{44}$
1018: ergs/s) and PKS 1622-253 ($L_{bol}\approx 2\times 10^{45}$ ergs/s)
1019: \citep{pun07,rod05}. All three have convincing evidence of a
1020: contemporaneous powerful jet and a weak (Seyfert 1 level) $L_{bol}$.
1021: The evidence for an active jet in PKSJ 1037-2705 is a powerful peak
1022: in the synchrotron SED, $\nu L_{\nu}>10^{45}$ ergs/s, in the IR
1023: based on archival 2MASS data. Similarly, the two epochs of X-ray
1024: data indicates that it is plausible that the inverse Compton peak of
1025: the SED in the X-ray band, (0.3 keV - 10 keV in the rest frame)
1026: exceeds $10^{45}$ ergs/s during the high states (ROSAT epoch). 3C
1027: 216 has a strong peak in the synchrotron SED $>10^{46}$ ergs/s. PKS
1028: 1622-253 is one of the strongest EGRET gamma ray sources, the
1029: gamma-ray apparent luminosity has a time average value of $\sim
1030: 10^{47}\mathrm{ergs/sec}$ and flares at $\sim
1031: 10^{48}\mathrm{ergs/sec}$ \citep{har99}. This circumstantial
1032: evidence tends to support the interpretation that even weak (Seyfert
1033: 1 level) accretion flows onto a black hole can produce a central
1034: engine for FRII jets. Thus, our study is supportive of the
1035: \citet{ogl06} interpretation of the weak Mid-IR, FRII NLRG
1036: subsample, they contain low luminosity nuclei.
1037: \par Finally, we comment on the conjecture in \citet{bor02,mac03}
1038: that these three Seyfert I galaxies, possessing a large jet power
1039: (at the FRII level), are not unusual objects, based on the small
1040: value of $L_{bol}/L_{Edd}$ in Seyfert 1 galaxies compared to QSOs
1041: \citep{sun89}. This conclusion is a natural consequence of the claim
1042: in \citet{bor02,mac03} that a small value of $L_{bol}/L_{Edd}$ is
1043: conducive to jet formation. The implication is that Seyfert I
1044: galaxies (with their relatively low values of $L_{bol}/L_{Edd}$
1045: compared to QSOs), possessing FR II radio power, should be a common
1046: state of AGN activity. However, this conjecture seems difficult to
1047: reconcile with the sparsity of known FR II level extended emission
1048: associated with broad line nuclei with Seyfert 1 level luminosity in
1049: optically selected samples (there is not a single example in the New
1050: General Catalog, Palomar-Green Survey or the Markarian catalog).
1051: Such a simple proposal is also at odds with the anecdotal case of
1052: PKS~0743$-$67 that was studied in detail in \citet{tin05} for just
1053: this reason. PKS~0743$-$67 is an example of a quasar that has an
1054: ultra-luminous accretion flow, $L_{bol}>2\times
1055: 10^{47}\mathrm{ergs/s}$, and has a very high Eddington rate,
1056: $L_{bol}/L_{Edd}\approx 1$. The jet kinetic luminosity was
1057: conservatively estimated at $\overline{Q}= 4.1 \times 10^{46}$
1058: ergs/sec which is 2.5 times that of Cygnus A. However, this was a
1059: conservative lower bound and the radio data also supports
1060: $\overline{Q}\approx 10^{47}$ ergs/sec, i.e one of the most powerful
1061: radio sources in the known Universe. Furthermore, PKS~0743$-$67 is
1062: presently active as evidenced by the powerful ($> 1 Jy$) unresolved
1063: VLBI radio core. A low Eddington ratio is not likely to be
1064: determinant to FRII jet production.
1065: \begin{acknowledgements}
1066: The discussion of optical data is based on observations made with
1067: the Italian Telescopio Nazionale Galileo (TNG) operated on the
1068: island of La Palma by the Fundación Galileo Galilei of the INAF
1069: (Istituto Nazionale di Astrofisica) at the Spanish Observatorio del
1070: Roque de los Muchachos of the Instituto de Astrofisica de Canarias.
1071: We are indebted to Matt Malkan and Ski Antonucci for valuable
1072: comments that help to estimate the uncertainties in the
1073: observational data. Tracy Clarke acknowledges that basic research in
1074: radio astronomy at the NRL is supported by 6.1 Base funding.
1075: \end{acknowledgements}
1076:
1077:
1078: \begin{thebibliography}{}
1079: \bibitem[Alloin et al (1988)]{all85} Alloin, D., Pelat, D., Phillips, M., Whittle, M, 1985,
1080: ApJ \textbf{288} 205
1081: \bibitem[Andernach et al(1988)]{and88} Andernach, H. et al, 1988,
1082: A\& AS \textbf{73} 265
1083: \bibitem[Antonucci (1988)]{ant88}Antonucci, R., 1988 Supermassive black holes in \emph{Proceedings of the Third George Mason Astrophysics Workshop,
1084: Fairfax, VA, Oct. 14-16, 1986} (Cambridge University Press, New
1085: York) p. 26-38
1086: \bibitem[Antonucci and Cohen(1983)]{ant83} Antonucci, R. and Cohen, R., 1983,
1087: ApJ \textbf{271} 564
1088: \bibitem[Antonucci and Ulvestad(1985)]{ant85} Antonucci, R. and Ulvestad, J., 1985,
1089: ApJ \textbf{294} 185
1090: \bibitem[Barthel et al(1990)]{bar90} Barthel, P., Tytler, D., Thompson, B. 1990, Astron. and Astrophys. Sup. \textbf{82} 339
1091: \bibitem[Best et al(2000a)]{bes01}Best, P., Röttgering, H., Lehnert, M. 2000 MNRAS \textbf{311} 1
1092: \bibitem[Best et al(2000b)]{bes00}Best, P., Röttgering, H., Lehnert, M. 2000 MNRAS \textbf{311}
1093: 23
1094: \bibitem[Blundell and Rawlings(2000)]{blu00} Blundell, K., Rawlings, S. 2000,
1095: AJ \textbf{119} 1111
1096: \bibitem[Boroson(2002)]{bor02}Boroson, T. 2002, ApJ
1097: \textbf{565} 78
1098: \bibitem[Bridle et al(1994)]{bri94} Bridle, A. et al 1994, AJ \textbf{108} 766
1099: \bibitem[Bridle and Perley (1984)]{bri84} Bridle, A. and Perley, R. 1984, Annu. Rev. Astron. Astrophys. \textbf{22}
1100: 319
1101: \bibitem[Celotti et al(1997)]{cel97}Celotti, A., Padovani and Ghisellini, G. 1997, MNRAS
1102: \textbf{286} 415
1103: \bibitem[Clarke(1999)]{cla99}Clarke, T.E. 1999, PhD
1104: Dissertation University of Toronto.
1105: \bibitem[Cleary et al.(2007)]{cle07} Cleary, K.; Lawrence, C. R.; Marshall, J. A.; Hao, L.; Meier, D 2007,
1106: ApJ \textbf{660} 117
1107: \bibitem[Colbert and Ptak(2002)]{col02}Colbert, E., Ptak, A. 2002, ApJS
1108: \textbf{143} 25
1109: \bibitem[Cutri et al (1985)]{cut85} Cutri, R., Wisniewski, W., Rieke, G., Lebofsky, M. 1985,
1110: ApJ \textbf{296} 423
1111: \bibitem[Elvis et al(1994)]{elv94} Elvis, M. et al 1994, ApJS
1112: \textbf{95} 1
1113: \bibitem[Fossati et al(1997)]{fos97}Fossati, G.; Celotti, A.;
1114: Ghisellini, G.; Maraschi, L 1997 MNRAS \textbf{289} 136
1115: \bibitem[Francis et al(1991)]{fra91}Francis, P. et al 1991, ApJ \textbf{373}
1116: 465
1117: \bibitem[Gutierrez(2006)]{gut06}Gutierrez, C. 2006 ApJL, \textbf{640}, 17
1118: \bibitem[Hartman et al (1999)]{har99}Hartman, R.. et al 1999, ApJS \textbf{123} 79
1119: \bibitem[Jakobsen et al (2003)]{jak03}Jakobsen, P., Jansen, R., Wagner, S., Reimers, D. 2003, A \& A \textbf{397}
1120: 891
1121: \bibitem [Kellermann et al (1969)]{kel69}Kellermann, K. I., Pauliny-Toth, I. I. K., Williams, P. J. S. 1969 ApJ \textbf{157} 1
1122: \bibitem[Kuehr, H et al.(1981)]{kuh81} Kuhr, H., Witzel, A., Pauliny-Toth, I.I.K.,
1123: Nauber, U. 1981, A \& AS \textbf{45}, 367
1124: \bibitem[Laor et al(1997)]{lao97} Laor, A. et al 1997, ApJ \textbf{477} 93
1125: \bibitem[Lind and Blandford(1985)]{lin85}Lind, K., Blandford, R.
1126: 1985, ApJ \textbf{295} 358
1127: \bibitem[Maccarone et al(2003)]{mac03} Maccarone, T., Gallo, E., Fender, R. 2003 MNRAS \textbf{345}, L19
1128: \bibitem[Ogle et al(2006)]{ogl06} Ogle, P., Whysong, D. and Antonucci, R.
1129: 2006 ApJ \textbf{647}, 161
1130: \bibitem[Punsly(1995)]{pun95}Punsly, B. 1995, AJ \textbf{109} 1555
1131: \bibitem[Punsly(2001)]{pun01}Punsly, B. 2001, \emph{Black Hole Gravitohydromagnetics} (Springer-Verlag, New York)
1132: \bibitem[Punsly(2005)]{pun05}Punsly, B. 2005, ApJL \textbf{623} 9
1133: \bibitem[Punsly(2006)]{pun07}Punsly, B. 2006, ApJL \textbf{651} 17
1134: \bibitem[Punsly and Tingay(2005)]{tin05}Punsly, B., Tingay, S. 2005, ApJL \textbf{633} L89
1135:
1136: \bibitem[Punsly et al (2005)]{rod05}Punsly, B., Rodriguez, L., Tingay, S., Cellone, S.
1137: 2005, ApJL \textbf{633} L93
1138: \bibitem[Rawlings and Saunders(1991)]{raw91} Rawlings, S., Saunders,
1139: R. 1991, Nature \textbf{349}, 138
1140: \bibitem[Schlegel et al.(1998)]{sch98} Schlegel, D. J., Finkbeiner, D. P. \& Davis, M. 1998, ApJ \textbf{500}
1141: 525
1142: \bibitem[Sun and Malkan(1989)]{sun89}Sun, W.H., Malkan, M. 1989, ApJ \textbf{346}
1143: 68
1144: \bibitem[Telfer et al(2002)]{tel02} Telfer, R., Zheng, W., Kriss, G.,
1145: Davidsen, A. 2002, ApJ \textbf{565} 773
1146: \bibitem[Vestergaard and Peterson(2006)]{ves06}Vestergaard, M. and Peterson, B. 2006, ApJ \textbf{641}
1147: 689
1148: \bibitem[Wang et al.(2004)]{wan04}Wang, J.-M., Luo, B, Ho, L. 2004, ApJL \textbf{615} 9
1149: \bibitem[Willott et al.(1999)]{wil99}Willott, C., Rawlings, S., Blundell, K., Lacy, M. 1999, MNRAS \textbf{309} 1017
1150: \bibitem[Zheng et al(1997)]{zhe97} Zheng, W. et al 1997, ApJ \textbf{475} 469
1151: \end{thebibliography}
1152:
1153:
1154:
1155: \end{document}
1156: