1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{graphics}
3: % To copy ApJ page style and fonts, uncomment next 2 lines:
4: %\usepackage{emulateapj5}
5: %\usepackage{apjfonts}
6:
7: \slugcomment{} \shorttitle{SDSS J104807.74+005543.5: Binary
8: Quasars?} \shortauthors{ZHOU ET AL.}
9:
10: \received{2003 November 29}
11: \begin{document}
12:
13: \title{Obscured Binary Quasar Cores in SDSS J104807.74+005543.5?}
14: \author{Hongyan Zhou, Tinggui Wang, Xueguang Zhang, Xiaobo Dong, and Cheng Li}
15: \email{mtzhou@ustc.edu.cn}
16:
17: \affil{Center for Astrophysics, University of Science and Technology of
18: China, Hefei, 230026, China}
19:
20:
21: \begin{abstract}
22: We report the discovery of a possible close binary system of
23: quasars in SDSS J1048+0055. The [OIII]$\lambda\lambda$4959,5007
24: emission lines are clearly double-peaked, and two discrete radio
25: sources with a projected physical separation of $\sim 20$ pc are
26: found in the VLBA milliarcsec resolution image at 8.4 GHz. Each of
27: the [O III]$\lambda \lambda 4959,5007$ doublets and H$\beta$ can
28: be well modelled by two Gaussians and the line ratio, $[O
29: III]\lambda 5007 /H\beta \sim 7$, is typical of Seyfert 2
30: galaxies. No broad component of H$\beta$ was detected and its $[O
31: III]\lambda 5007$ luminosity, $L_{[OIII]}\approx 9.2\times
32: 10^{42}$ erg $s^{-1}$, is comparable to luminous quasars and is a
33: few ten times more luminous than typical Seyfert galaxies. One
34: natural interpretation is that SDSS J1048+0055 contains two close
35: quasar-like nuclei and the BLR around them are obscured. Other
36: possible models are also discussed. We suggest that double-peaked
37: narrow emission line profile may be an effective way of selecting
38: candidates of binary black holes with intermediate separation.
39: \end{abstract}
40:
41: \keywords{galaxies: active --- galaxies: individual (SDSS J1048+0055) --- quasars: emission lines --- radio continuum: galaxies --- black hole physics}
42:
43: \section{Introduction}
44:
45: Black holes are the simplest objects in the universe and merging
46: of Binary Black Holes (BBH) can produce strong gravitational
47: radiation. Hence, BBH systems are ideal laboratories for exploring
48: and testing strong field relativistic physics. There is a growing
49: body of evidence that black holes exist on many mass scales, from
50: stellar mass, up to $\sim $ a few $10^{9}$$M_{\odot }$. Numerous
51: clues, especially those provided by the Hubble Space Telescope,
52: suggest that almost every galaxy harbors a Supermassive Black Hole
53: (SMBH) in its center (e.g., Sarzi et al. 2001; Pinkney et al.
54: 2003). According to the hierarchical merger scenario of galaxy
55: evolution, coalescence is one of the most important processes in
56: galaxy evolution and a typical galaxy may experience between one
57: and several such episodes during its lifetime (Struck 1999).
58: Merger of galaxies can bring two SMBHs together eventually to form
59: a BBH, which can survive for about one Hubble time in spherical
60: galaxies, axisymmetric galaxies or certain triaxial galaxies (Yu
61: 2002). Therefore the presence of two SMBHs in one galaxy forming
62: supermassive BBH, should not be a rare phenomenon.
63:
64: Enormous efforts have been made in recent years to find
65: observational evidence for the existence of supermassive BBH.
66: Evidence is sought in two directions: spatially resolved systems
67: where the BBH can be identified morphologically and unresolved
68: systems where the existence of the BBH can be inferred in various
69: ways (see Komossa 2003 for an extensive review). Examples of the
70: latter are as follows. Winged or X-shaped radio sources were
71: suggested to be BBH merger remnants (e.g., Merritt \& Ekers 2002;
72: Wang et al. 2003; Liu, Wu, \& Cao 2003). Helically distorted radio
73: jets observed in some objects were interpreted sometimes as due to
74: the orbital motion or precession of the black hole producing the
75: jets (e.g., Sudou et al. 2003). Some BBH models predict the
76: surface brightness of the host galaxies would decrease toward the
77: galactic center and such an effect has indeed been observed in a
78: few early-type galaxies (e.g. Lauer et al. 2002). Rather more
79: direct evidence comes from resolved systems: several pairs of
80: quasars with almost the same redshift and angular separations
81: between $\sim 3-10"$ have been observed and were taken to be real
82: binary quasars. The projected physical distance between these
83: potential binary quasars is usually a few ten kpc (Mortlock,
84: Webster, \& Francis 1999). NGC 6240 is perhaps the only case where
85: binary active nuclei are separately identified: the high
86: resolution observation of Chandra revealed two hard X-ray nuclei
87: in this interesting object and the projected physical separation
88: of about 1.4 kpc (Komossa et al. 2003), which is the smallest
89: among known binary AGN. As far as we know, no quasar pair with
90: physical separation $\lesssim 1$ kpc has been yet found.
91:
92: Of the above lines of evidence for the existence of supermassive
93: BBH, binary AGN is of particular interest. It must, however, be
94: admitted that apart from NGC 6240, the evidence so far cannot be
95: considered as conclusive and efforts must be continued to seek out
96: suppermassive BBH. In this letter, we report the discovery of a
97: new binary quasar candidate, SDSS J104807.74+005543.5 (hereafter
98: SDSS J1048+0055). It is identified from the Sloan Digital Sky
99: Survey Data Release 1 (SDSS DR1, Abazajian et al. 2003) among
100: $\sim 100$ low redshift galaxies that show double-peaked narrow
101: emission lines, most of which are active. Apart from the presence
102: of such lines, the other piece of evidence that sustains the BBH
103: interpretation is that its high angular resolution VLBA image
104: reveals two discrete radio sources, possibly corresponding to two
105: different active nuclei. The [OIII] luminosity of the two nuclei
106: are in the range for quasars. The discovery of SDSS J1048+0055 and
107: more than one hundred of double-peaked narrow line objects in the
108: spectral dataset of the SDSS DR1 might suggest that this be an
109: effective way of selecting candidates of binary black holes with
110: intermediate, sub-kpc separations. We postpone detailed analyse of
111: the whole sample of double-peaked narrow emission lines to a
112: forthcoming paper. Throughout this letter, an $H_{0}=$70 km~s$^{-1}$~Mpc$%
113: ^{-1}$, $\Omega_{M}=0.3$, and $\Omega_{\Lambda} =0.7$ cosmology is
114: assumed.
115:
116: \section{Data analysis}
117:
118: \subsection{Optical spectrum}
119:
120: SDSS J1048+0055 is observed by the SDSS only because it is the
121: counterpart of a FIRST radio source and classified as 'QSO' by the
122: spectroscopic pipeline of the SDSS (Stoughton et al. 2002). %Its
123: %optical spectrum in the observer's frame is shown in Figure 1 with
124: %recognizable emission lines labelled.
125: It falls in the DR1 quasar
126: catalog (Schneider et al. 2003) due to its high luminosity
127: ($M_{i}=-22^{m}.63$) and seemingly broad line width.
128:
129: The optical continuum of SDSS J1048+0055, which is rather noisy,
130: is fitted with a power law. The fitted continuum is subtracted
131: before measurement of the emission lines. Because the
132: [OIII]$\lambda \lambda $4959, 5007 doublets clearly show a
133: double-peaked profile and a rather broad wing, the continuum
134: subtracted spectrum in the H$\beta
135: +$ [OIII]$\lambda \lambda $4959, 5007 regime is fitted with $2\times
136: 3+1$ Gaussian components, i.e., two sets of three Gaussians for
137: the two peaks and the broad wing of the [OIII]$\lambda \lambda
138: $4959,5007 doublets and one Gaussian for H$\beta $. Within each
139: set of the components for [OIII], the centroids of the three
140: Gaussians are forced to correspond to a same redshift; their
141: widths to a same Doppler broadening; also the intensity ratio of
142: [OIII]$\lambda $5007 to [OIII]$\lambda $4959 is fixed at its
143: theoretical value. The final fit is done through minimization of
144: $\chi ^{2}$ and the result is shown in Figure 1 upper panel. The
145: lower panel displays the [OII]$\lambda $3727 line, which is the
146: 2nd strongest after [OIII]$\lambda $5007. However, the [OII] line
147: is single-peaked and can be satisfactorily modelled by a single
148: Gaussian.
149:
150: We would like to point out that H$\beta$ might also be double-peaked.
151: When fitted with a single Gaussian, its line width $FWHM(H\beta )\approx
152: 1200 km s^{-1}$ is much broader than [OII] and each of the [OIII] peaks.
153: The line can be fitted, comparably, with the template of [OIII] line
154: profile, Other high
155: ionization narrow lines, such as [NeIII]$\lambda $3869 and [NeIII]
156: $\lambda $3967 also seem to show double-peaked profile but higher
157: quality spectral observation is needed to confirm this feature. We
158: did not analyse other emission lines because the S/N ratio is not
159: high enough for us to draw any significant conclusion. The
160: measured line parameters are assembled in Table 1.
161: \clearpage
162: \tablenum{1}
163: \begin{deluxetable}{lllll}
164: \tablecaption{Line Parameters of SDSS J1048+0055.\label{tbl-1}}
165: %\rotate
166: \tablewidth{0pt} \tablehead{
167: \colhead{Line\tablenotemark{a}} & \colhead{Center Wavelength\tablenotemark{b}} & redshift & \colhead{Flux\tablenotemark{b}} & \colhead{FWHM} \\
168: \colhead{} & \colhead{$\mathring{A}$} & & \colhead{10$^{-17}$erg
169: s$^{-1}$ cm$^{-2}$} & \colhead{km s$^{-1}$} } \startdata
170: $[OIII]\lambda 5007_{r}$ & $8238.5\pm 0.3$ & 0.6450 & $161\pm 11$ & $553\pm 25$ \\
171: $[OIII]\lambda $5007$_{b}$ & $8218.5\pm 0.5$ & 0.6410 & $209\pm 13$ & $610\pm 24$ \\
172: $[OIII]\lambda $5007$_{br}$ & $8224.7\pm 0.9$ & 0.6423 & $151\pm 11$ & $2573\pm 198$ \\
173: $[OIII]\lambda $4959$_{r}$ & 8159.5 & 0.6450 & 52 & 553 \\
174: $[OIII]\lambda $4959$_{b}$ & 8139.7 & 0.6410 & 67 & 610 \\
175: $[OIII]\lambda $4959$_{br}$ & 8142.5 & 0.6423 & 49 & 2573 \\
176: $H\beta $ & $7987.5\pm 1.0$ & 0.6422 & $76\pm 6$ & $1238\pm 102$ \\
177: $[OII]\lambda $3727 & $6122.5\pm 0.2$ & 0.6422 & $135\pm 9$ & $721\pm 36$ \\
178: $MgII\lambda $2800 & $4600.1\pm 1.0$ & 0.6430 & $98\pm 10$ &
179: $1806\pm 141$
180:
181: \enddata
182: \tablenotetext{a}{'r', 'b' and 'br' denote red, blue and broad
183: component of the $[OIII]\lambda \lambda 4959,5007$ emission lines,
184: respectively. } \tablenotetext{b}{The central wavelengths are in the
185: observer's frame.}
186: %and the line widths (FWHM, full width at half-maximum
187: %intensity) are values in the rest frame. }
188:
189: \end{deluxetable}
190: \clearpage
191: \subsection{Radio properties}
192: SDSS J1048+0055 was observed many times in several radio
193: band-passes, from 0.365 to 8.4 GHz. It was observed at 1.4 GHz
194: during NVSS in 1995-02 (Condon et al. 1998) and during FIRST in
195: 1998-08 (White et al. 1997). There is no significant difference in
196: the flux observed on these two occasions ($S_{\nu,NVSS}=270.9\pm
197: 8.1 $ mJy and $S_{\nu,FIRST}=270.7$ mJy). This flux is also
198: marginally consistent with an early measurement in the Green Bank
199: 1.4 GHz Northern Sky Survey, which is $S_{\nu,1.4 GHz}\approx 299$
200: with uncertainty of $\sim 25-30$ mJy (White \& Becker 1992).
201:
202: However, the 4.85 GHz flux of $340\pm 21$ mJy given in the
203: MIT-Green Bank 5 GHz Survey Catalog (MGBS, Bennett et al. 1986) is
204: different (at $\sim 2\sigma $ level) from that presented by
205: Gregory \& Condon 1991 ($266\pm 37$ mJy), Becker, White, \&
206: Edwards 1991 ($262\pm 39$ mJy) and Griffith et al. 1995 ($262\pm
207: 39$ mJy). The flux difference between the MGBS and the later
208: surveys cannot be explained by the larger beamsize of MGBS
209: (FWHP$=2'.8$): 1) because the source is compact, and 2) because
210: within $5'$ of the source NVSS detected no other sources while
211: FIRST only detected two with insignificant fluxes around 1 mJy. It
212: seems likely that this variation with an amplitude $\sim 39\% $ is
213: real. This interpretation is, moreover, consistent with its
214: compact morphology and its flat spectrum.
215:
216: The overall radio spectrum is moderately flat with a spectral
217: slope of $\alpha _{\nu ,0.365-8.4GHz}=0.13\pm 0.03$ (S$_{\nu
218: }\varpropto \nu ^{-\alpha }$). The radio luminosity between
219: $0.365-8.4$ GHz, $ L_{0.365-8.4GHz}\approx 4.2\times 10^{43}$ erg
220: s$^{-1}$, is quite high even for a radio-loud quasar. The flatness
221: of the spectrum is consistent with the compact radio morphology.
222: It is unresolved by FIRST at angular resolution $\sim$ 5''.4. This
223: object is also observed by the VLBA Calibrator Survey (VCS1,
224: Beasley et al. 2002) and remains unresolved at 2.3 GHz with
225: resolution ellipse of $\sim 5\times10$ milliarcsecs. However, the
226: higher resolution 8.4 GHz radio image ($\sim 1.5\times 3$
227: milliarcsecs) clearly shows two compact sources with the west
228: source $\sim$ 30 times stronger than the east one (Figure 2). The
229: angular offset between the two sources is about 2.5 milliarcsecs,
230: corresponding to a projected physical distance of $\sim 20$ pc at
231: a redshift of 0.6422 derived from the [OII] and $H\beta $ emission
232: line. The large difference in the radio flux, the small angular
233: separation and flat spectrum conspire to indicate that we are dealing
234: with possible two separate radio sources, rather than with two lobes
235: of a single source. Further measurement of the radio spectrum of the
236: weak component would be necessary to confirm this.
237:
238: \section{Discussion and future prospect}
239:
240: \subsection{SDSS J1048+0055 as obscured binary quasars}
241:
242: A straightforward interpretation of the double-peaked profile of [OIII]$%
243: \lambda \lambda $4959,5007 is that they come from two distinct narrow
244: emission line regions (NLR) around two active nuclei. The [OIII]$%
245: \lambda $5007 luminosity of the red and blue peaks,
246: $L_{[OIII],red}\approx 2.8\times 10^{42}$ erg s$^{-1}$ and
247: $L_{[OIII],blue}\approx 3.6\times 10^{42} $ erg s$^{-1}$, is
248: comparable to that of luminous quasars in the BQS sample, and at
249: least 10 times more luminous than Seyfert galaxies. SDSS
250: J1048+0055 is rather faint in the optical in comparison with the
251: radio. The flux ratio of radio to optical or the Radio-Loudness,
252: $RL\equiv f_{r}/f_{o}\sim 1.5-2.1\times 10^{4}$, is
253: extraordinarily large. Such a large RL implies that the nuclear
254: emission in the optical is heavily obscured. The radio morphology
255: also suggests that two active nuclei may be present in the center
256: of SDSS J1048+0055 as was pointed out in the last section.
257: If the two discrete radio sources revealed by VLBA proved to be
258: stationary and if the weak component also shows flat spectrum by
259: further VLBI observations, thereby thus confirming
260: their BBH status, then we can say that the separation between the
261: BBH should not be much larger than tens of pc. Then SDSS J1048+0055
262: would be the only galaxy harboring BBH with such a small separation
263: and the system may be at the last stage before the two nuclei (BHs) become
264: a bound system (Yu 2002). The fact that [OIII]$\lambda \lambda 4959,5007$
265: show double-peaked profile while [OII]$\lambda 3727$ does not is
266: understandable because high ionization emission line regions are often
267: more compact than low ionization ones. It is likely that double-peaked
268: [OIII] emission line is mainly concentrated in the two cores of NLR
269: whose centers coincide with the two radio peaks while the narrower and single-peak
270: [OII] may be emitted by the whole NLR. Therefore SDSS J1048+0055 would
271: most likely be a galaxy merger with two quasar cores in its central
272: region\footnote{Readers may find other lines of evidence that support
273: the BBH interpretation in the detailed discussion of the whole sample of
274: double-peaked narrow emission line from SDSS DR1 (Wang et al. 2004, in preparation)}.
275:
276:
277: \subsection{The efficiency of finding BBH using double-peaked narrow
278: emission lines}
279:
280:
281: Now that most galaxies harbor a supermassive black hole and the
282: frequency of galaxy collisions is relatively high, closely bound
283: BBH in a single galaxy should not be a rare occurrence.
284: Theoretical calculations show that coalescence of two black hole
285: goes through several stages. Dynamic friction causes the black
286: holes along with their surrounding stellar cluster to approach
287: each other. As the stars in the cluster being stripped off, the
288: the two black holes form a BBH. Some BBH will further lose angular
289: momentum through the scattering of passing-by stars, but analysis
290: shows that some BBH can survive for about one Hubble time in
291: large, spherical galaxies and the orbital velocities of the
292: surviving BBH are generally believed to be $\sim 10^{2-4}$ km
293: s$^{-1}$ (Yu 2002; c.f., Milosavljevi{\' c} \& Merritt 2003).
294:
295: Detection of supermassive BBH and the estimation of the frequency
296: of such events can provide important constraints on models of
297: galaxy formation and evolution. Now, the evolution timescale of
298: BBH is extremely uneven. A sharp rise occurs at $\sim 10$ pc and
299: the timescales for a separation of $\lesssim $ a few pc are
300: several orders of magnitude greater than for $\sim 10^{1-4}$ pc
301: (Yu 2002). The time it takes a BBH of similar masses to evolve
302: from a few kpc to $\sim 10$ pc is $\sim 10^{8} $ yr, and this time
303: increase with increasing mass ratio of the BBH. Therefore, even
304: every galaxy experiences a major merger during its lifetime and
305: most of BBH survive, the absolute majority would have separation
306: $\lesssim $ a few pc, which is beyond the resolution limit of most
307: of present instruments. While double nuclei with separation of
308: several kpc or larger have been identified in some galaxies,
309: however, BBH with separation $\sim 10^{1-3} $pc have not been
310: found yet.
311:
312: It is likely that merging will trigger the activity of both black
313: holes at intermediate separations. Gas concentration and starburst
314: activity in the nuclei were observed in interacting galaxies at
315: even large separation (e.g., Gao \& Solomon 2003), and a pair of
316: binary AGN are detected in NGC 6240. The associated star-cluster
317: or bulge around each black hole may survive in a fraction of such
318: systems at this distance. Thus each nucleus may well possess its
319: own NLR. In this case, our calculations show that about 5-10\% of
320: such systems will be found as double-peaked sources under the
321: generic assumption of virialization of NLR. If a significant
322: fraction of AGN are indeed in such merging systems, a few percent
323: of AGN should show double-peaked line profile. We found $\sim$ 100
324: objects that shows this feature in at least one narrow line from
325: the subsample of $\sim 10^4$ SDSS galaxies and quasars that either
326: [OIII] or H$\alpha$ have S/N ratio $\gtrsim 30 \sigma $. The
327: fraction of such objects is about 1\%, and increases with the
328: level of nuclear activity. Hence, the simple model estimation is
329: quite consistent with the result of observation. The model and the
330: sample will be described in detail elsewhere.
331:
332: \subsection{Other possible explanations and future prospect}
333:
334: We would like to point out that the BBH interpretation for SDSS
335: J1048+0055, as well as for the other objects in our double-peaked
336: narrow line sample, is certainly not the only one. Two other
337: possible interpretations are bipolar outflow and a disk-like
338: emission line region. The latter seems unlikely because the
339: compact structure and flat spectrum of SDSS J1048+0055 suggest
340: that the system is observed face-on, while an inclined disk is
341: required to explain the double peaked profile in SDSS J1048+0055.
342: On the other hand, if the bipolar outflow interpretation is
343: correct, then we can only explain the weaker radio source as a
344: one-sided jet like that of 2255-282 which is an optical variable
345: radio source. For the latter, a jet is revealed by VLBA in the 5
346: GHz image and its 15 GHz images show radio structure similar to
347: that of the former (Zensus et al. 2002 and references therein).
348: The VLBA radio image of SDSS J1048+0055 was taken some 8 years
349: ago, hence, another high angular resolution radio image can tell
350: whether this is indeed the case.
351:
352: About $25\%$ of the objects in our double-peaked narrow line
353: sample are detected in the radio by the FIRST and $\sim 10$
354: objects, including SDSS J1048+0055 and SDSS J094144.82+575123.7,
355: have radio flux $\gtrsim 10$ mJy. According to the estimates made
356: in \S 3.3, some of these objects should be resolved into two
357: discrete radio sources at $\sim $ milliarcsec resolution if the
358: BBH interpretation is indeed correct. A sizable fraction of the
359: objects in our sample may also be resolved in the optical by HST.
360:
361: \acknowledgements We thank Prof. T. Kiang for help in English. We are
362: grateful to Drs Youjun Lu \& Qingjuan Yu for enlightened discussion.
363: This work is supported
364: by the Chinese National Science foundation (NSF 19925313 and
365: 10233030) and the key program of Chinese ministry of science and
366: technology. This paper has made use the data from NED, NRAO, and
367: SDSS.
368:
369: Funding for the creation and distribution of the SDSS Archive has been
370: provided by the Alfred P. Sloan Foundation, the Participating Institutions,
371: the National Aeronautics and Space Administration, the National Science
372: Foundation, the U.S. Department of Energy, the Japanese Monbukagakusho, and
373: the Max Planck Society. The SDSS Web site is http://www.sdss.org/.
374:
375: The SDSS is managed by the Astrophysical Research Consortium (ARC) for the
376: Participating Institutions. The Participating Institutions are The
377: University of Chicago, Fermilab, the Institute for Advanced Study, the Japan
378: Participation Group, The Johns Hopkins University, Los Alamos National
379: Laboratory, the Max-Planck-Institute for Astronomy (MPIA), the
380: Max-Planck-Institute for Astrophysics (MPA), New Mexico State University,
381: Princeton University, the United States Naval Observatory, and the
382: University of Washington.
383:
384: \begin{thebibliography}{}
385:
386: \bibitem[Abazajian et al.(2003)]{2003AJ....126.2081A} Abazajian, K.~et al.\ 2003, \aj, 126, 2081
387: \bibitem[Beasley et al.(2002)]{2002ApJS..141...13B} Beasley, A.~J., Gordon, D., Peck, A.~B., Petrov, L., MacMillan, D.~S., Fomalont, E.~B., \& Ma, C.\ 2002, \apjs, 141, 13
388: \bibitem[Becker, White, \& Edwards(1991)]{1991ApJS...75....1B} Becker, R.~H., White, R.~L., \& Edwards, A.~L.\ 1991, \apjs, 75, 1
389: \bibitem[Bennett et al.(1986)]{1986ApJS...61....1B} Bennett, C.~L., Lawrence, C.~R., Burke, B.~F., Hewitt, J.~N., \& Mahoney, J.\ 1986, \apjs, 61, 1
390: \bibitem[Condon et al.(1998)]{1998AJ....115.1693C} Condon, J.~J., Cotton, W.~D., Greisen, E.~W., Yin, Q.~F., Perley, R.~A., Taylor, G.~B., \& Broderick, J.~J.\ 1998, \aj, 115, 1693
391: \bibitem[Gao \& Solomon(2003)]{418} Gao. Y \& Solomon P. M., astro-ph/0310339
392: \bibitem[Gregory \& Condon(1991)]{1991ApJS...75.1011G} Gregory, P.~C.~\& Condon, J.~J.\ 1991, \apjs, 75, 1011
393: \bibitem[Griffith, Wright, Burke, \& Ekers(1995)]{1995ApJS...97..347G} Griffith, M.~R., Wright, A.~E., Burke, B.~F., \& Ekers, R.~D.\ 1995, \apjs, 97, 347
394: \bibitem[Komossa (2003)]{421} Komossa, S., astro-ph/0306439
395: \bibitem[Komossa et al.(2003)]{2003ApJ...582L..15K} Komossa, S., Burwitz, V., Hasinger, G., Predehl, P., Kaastra, J.~S., \& Ikebe, Y.\ 2003, \apjl, 582, L15
396: \bibitem[Lauer et al.(2002)]{2002AJ....124.1975L} Lauer, T.~R.~et al.\ 2002, \aj, 124, 1975
397: \bibitem[Liu, Wu, \& Cao(2003)]{2003MNRAS.340..411L} Liu, F.~K., Wu, X., \& Cao, S.~L.\ 2003, \mnras, 340, 411
398: \bibitem[Merritt \& Ekers(2002)]{2002Sci...297.1310M} Merritt, D.~\& Ekers, R.~D.\ 2002, Science, 297, 1310
399: \bibitem[Milosavljevi{\' c} \& Merritt(2003)]{2003ApJ...596..860M} Milosavljevi{\' c}, M.~\& Merritt, D.\ 2003, \apj, 596, 860
400: \bibitem[Mortlock, Webster, \& Francis(1999)]{1999MNRAS.309..836M} Mortlock, D.~J., Webster, R.~L., \& Francis, P.~J.\ 1999, \mnras, 309, 836
401: \bibitem[Pinkney et al.(2003)]{2003ApJ...596..903P} Pinkney, J.~et al.\ 2003, \apj, 596, 903
402: \bibitem[Sarzi et al.(2001)]{2001ApJ...550...65S} Sarzi, M., Rix, H., Shields, J.~C., Rudnick, G., Ho, L.~C., McIntosh, D.~H., Filippenko, A.~V., \& Sargent, W.~L.~W.\ 2001, \apj, 550, 65
403: \bibitem[Schneider et al.(2003)]{2003AJ....126.2579S} Schneider, D.~P.~et al.\ 2003, \aj, 126, 2579
404: \bibitem[Stoughton et al.(2002)]{2002AJ....123..485S} Stoughton, C.~et al.\ 2002, \aj, 123, 485
405: \bibitem[Struck(1999)]{1999PhR...321....1S} Struck, C.\ 1999, \physrep, 321, 1
406: \bibitem[Sudou, Iguchi, Murata, \& Taniguchi(2003)]{2003Sci...300.1263S} Sudou, H., Iguchi, S., Murata, Y., \& Taniguchi, Y.\ 2003, Science, 300, 1263
407: \bibitem[Wang, Zhou, \& Dong(2003)]{2003AJ....126..113W} Wang, T., Zhou, H., \& Dong, X.\ 2003, \aj, 126, 113
408: \bibitem[White \& Becker(1992)]{1992ApJS...79..331W} White, R.~L.~\& Becker, R.~H.\ 1992, \apjs, 79, 331
409: \bibitem[White, Becker, Helfand, \& Gregg(1997)]{1997ApJ...475..479W} White, R.~L., Becker, R.~H., Helfand, D.~J., \& Gregg, M.~D.\ 1997, \apj, 475, 479
410: \bibitem[Yu(2002)]{2002MNRAS.331..935Y} Yu, Q.\ 2002, \mnras, 331, 935
411: \bibitem[Zensus et al.(2002)]{2002AJ....124..662Z} Zensus, J.~A., Ros, E., Kellermann, K.~I., Cohen, M.~H., Vermeulen, R.~C., \& Kadler, M.\ 2002, \aj, 124, 662
412:
413: \end{thebibliography}
414: \clearpage
415:
416:
417:
418: \begin{figure}[tbp]
419: \epsscale{0.8} \plotone{f1.eps} \caption{Gaussian fitting of
420: H$\protect\beta $+ [OIII] regime (upper panel) and [OII] (lower
421: panel) after continuum substraction. Dotted lines denote the fit
422: residuals.} \label{fig-1}
423: \end{figure}
424:
425:
426: \begin{figure}[tbp]
427: \epsscale{1.0}
428: \plotone{f2.eps} \caption{The 8.4-GHz (left panel)
429: and 2.3-GHz (right panel) radio morphology of SDSS J1048+0055 from
430: the VCS1 survey. Contour levels are -1,1,2,4,8, etc. times the
431: lowest contour level. SDSS J1048+0055 is only marginally resolved
432: at 2.3-GHz, but higher resolution 8.4-GHz image clearly reveals
433: two discrete radio sources. }
434: \label{fig-2}
435: \end{figure}
436:
437: \end{document}
438: