1: %\documentstyle[referee,epsf]{mn}
2: \documentstyle[epsf]{mn}
3:
4: %\documentclass{mn2e}
5: \hyphenation{english}
6: %%\input psfig.tex
7:
8:
9: \newif\ifAMStwofonts
10: %\AMStwofontstrue
11:
12:
13: %%%%% AUTHORS - PLACE YOUR OWN MACROS HERE %%%%%
14:
15:
16: \def\gtorder{\mathrel{\raise.3ex\hbox{$>$}\mkern-14mu
17: \lower0.6ex\hbox{$\sim$}}}
18: \def\ltorder{\mathrel{\raise.3ex\hbox{$<$}\mkern-14mu
19: \lower0.6ex\hbox{$\sim$}}}
20:
21: %\newcommand\ion[2]{#1$\;${\small\rmfamily\@Roman{#2}}\relax}
22:
23:
24: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
25:
26:
27:
28:
29: %\slugcomment{Draft of \today}
30:
31: %\shorttitle{SDSS~J$131339.98+515128.3$}
32:
33: %\shortauthors{Ofek et al.}
34:
35:
36: \title[SDSS J131339.98$+$515128.3]{SDSS J131339.98$+$515128.3: A new gravitationally lensed quasar selected based on near-infrared excess}
37:
38: \author[Ofek et al.]
39: {Eran~O.~Ofek\thanks{e-mail: eran@astro.caltech.edu}$^{1}$, Masamune~Oguri$^{2}$, Neal~Jackson$^{3}$, Naohisa~Inada$^{4}$, and Issha~Kayo$^{5}$ \\
40: $^{1}$ Division of Physics, Mathematics and Astronomy, California Institute of Technology, Pasadena, CA 91125 \\
41: $^{2}$ Kavli Institute for Particle Astrophysics and Cosmology, Stanford University, Menlo Park, CA 94025. \\
42: $^{3}$ The University of Manchester, Jodrell Bank Observatory, Macclesfield, Cheshire SK11 9DL, UK. \\
43: $^{4}$ Cosmic Radiation Laboratory, RIKEN (The Institute of Physical and Chemical Research), 2-1 Hirosawa, Wako, Saitama 351-0198, Japan. \\
44: $^{5}$ Department of Physics and Astrophysics, Nagoya University, Chikusa-ku, Nagoya 464-8602, Japan}
45:
46:
47: \date{Accepted ?
48: Received ?
49: in original form ?}
50:
51:
52: \begin{document}
53:
54: \maketitle
55:
56: \begin{abstract}
57:
58: We report the discovery of a new gravitationally lensed quasar, SDSS
59: J131339.98$+$515128.3, at a redshift of $1.875$ with an image separation
60: of $1\farcs24$. The lensing galaxy is clearly detected in visible-light
61: follow-up observations. We also identify three absorption-line doublets in
62: the spectra of the lensed quasar images, from which we measure the lens
63: redshift to be $0.194$. Like several other known lenses, the lensed
64: quasar images have different continuum slopes. This difference is
65: probably the result of reddening and microlensing in the lensing galaxy.
66: The lensed quasar was selected by correlating Sloan Digital Sky Survey
67: (SDSS) spectroscopic quasars with Two Micron All Sky Survey (2MASS)
68: sources and choosing quasars that show near-infrared (IR) excess.
69: The near-IR excess can originate, for example, from the contribution of
70: the lensing galaxy at near-IR wavelengths. We show that the near-IR
71: excess technique is indeed an efficient method to identify lensed
72: systems from a large sample of quasars.
73:
74: \end{abstract}
75:
76: \begin{keywords}
77: gravitational lensing\ --- quasars: individual (SDSS J131339.98$+$515128.3)
78: \end{keywords}
79:
80:
81: \section{Introduction}
82: \label{Introduction}
83:
84: Construction of large samples of gravitational lenses has been shown
85: to be a unique tool for studying galaxy mass profiles
86: (e.g., Cohn et al. 2001; Rusin et al. 2002;
87: Koopmans \& Treu 2003; Rusin \& Kochanek 2005; Treu et al. 2006),
88: galaxy evolution (e.g., Kochanek et al. 2000; Ofek et
89: al. 2003; Treu \& Koopmans 2004; Rusin \& Kochanek 2005), and cosmology
90: (e.g., Linder 2004; Oguri 2007). Moreover, individual lenses enable us
91: to study the lensed quasar engine (e.g., Kochanek et al. 2007) and the
92: dark matter content of galaxies through microlensing (e.g., Wambsganss
93: et al. 2000; Schechter \& Wambsganss 2002).
94:
95: The photometric and spectroscopic quasar samples of
96: the Sloan Digital Sky Survey (SDSS; York et al. 2000) were estimated to
97: contain $\sim1000$ lensed quasars. However, in practice, because of
98: seeing limitations, it is challenging to find most of these lenses. The
99: most successful search for lensed quasars in the SDSS (Oguri et
100: al. 2006) utilizes the morphological properties of SDSS quasars, as well
101: as colours of nearby objects, to efficiently select lensed quasars with
102: image separation greater than $1''$ and flux ratio above
103: $\sim0.3$. Lensed quasars can also be discovered spectroscopically by
104: looking for a compound spectrum consisting of a galaxy and a quasar
105: (e.g., Johnston 2003).
106: %A high-resolution snapshot survey of high-redshift ($z>4$)
107: %quasars discovered no gravitationally lensed quasar (Richards et
108: %al. 2004, 2006).
109: Since these techniques locate only a fraction of lensed
110: quasars in the SDSS, it is of great importance to develop additional
111: approaches in order to find more lenses, especially lenses with image
112: separation below $\sim1''$.
113:
114: The efficiency of lens surveys can be enhanced by combining
115: multi-wavelength data. For instance, Kochanek et al. (1999) compared
116: optical and radio properties of wide-separation quasar pairs to argue
117: that they are binary quasars rather than gravitational lenses.
118: %
119: Jackson \& Browne (2007) proposed a lens search method that exploits the
120: small position differences between radio positions of lensed images and
121: optical positions of lensing galaxies. The method may
122: enhance the efficiency of lens searches by up to a factor of $\sim 10$.
123: %
124: Another approach, presented by Haarsma et al. (2005),
125: is to look for extended radio lobes which are
126: lensed by galaxies along the line of sight.
127: %
128: We note that the magnification bias for such multi-wavelength
129: surveys is more complicated than usual
130: (e.g., Borgeest et al. 1991; Ofek et al. 2002; Wyithe et al. 2003).
131:
132:
133: In this paper we present an approach for finding lensed quasars in
134: the SDSS data. This method relies on selecting, as candidate lens
135: systems, quasars with near-IR excess, which may be due to the
136: contribution of a lensing galaxy.
137: %compares SDSS quasars with near-infrared (IR)
138: %data to select near-IR quasars as lens candidates. This is because a
139: %typical lens system consists of blue quasar images and a red lensing
140: %galaxy, and therefore at near-IR regime the lensing galaxy tends to
141: %dominate the flux of the system.
142: We note in this context that Gregg et al. (2002) suggested that red
143: quasar samples may contain a large number of lensed quasars due to
144: reddening by the lensing galaxy. We show that the quasar
145: SDSS~J$131339.98+515128.3$ selected by this method is a
146: new gravitationally lensed quasar, which is confirmed from imaging and
147: spectroscopic follow-up observations.
148:
149: The outline of this paper is as follows. We describe the near-IR excess
150: selection method and the lensed quasar candidates in \S\ref{Method}. In
151: \S\ref{Obs} we present the observations of the new lens we have found,
152: and its spectroscopy and photometry are described in \S\ref{Phot}. We
153: model the new system in \S\ref{Model} and discuss the results in
154: \S\ref{Disc}. Finally the conclusions are given in \S\ref{Conc}.
155:
156:
157: \section{Selection of lensed quasar candidates based on the near-IR excess method}
158: \label{Method}
159:
160:
161: The selection process for gravitational lenses presented here is based
162: on the fact that most lenses are red galaxies whereas quasars are
163: much bluer. Therefore, even if the flux from the lens galaxy does not
164: dominate at visible-light frequencies, which is quite common among lens
165: systems discovered to date, it is expected that in some cases the
166: lensing galaxy has a non-negligible contribution to the integrated
167: brightness of the lens system at near-IR frequencies. This suggests that
168: some of the lens systems may be characterized by a near-IR excess
169: that originates from the lensing galaxy. Figure~\ref{Example_GalQSO_Spec}
170: illustrates this idea by showing template spectra of quasars and
171: early-type galaxies redshifted to typical source and lens redshifts, respectively.
172: %
173: \begin{figure}
174: \centerline{\epsfxsize=8.5cm\epsfbox{f1.eps}}
175: \caption{Illustration of the spectra of an elliptical galaxy (Mannucci
176: et al. 2001) at a redshift $0.5$ and a quasar (Vanden Berk et al. 2001)
177: at a redshift of $2.0$. The spectra are normalized such that the
178: $i$-band magnitude difference is $i_{\rm gal}-i_{\rm QSO}=1$, which is
179: typical for lensed quasar systems. The transmission curves of the SDSS
180: $ugriz$ and the 2MASS $JHK_{s}$ filters are also shown (Moro \& Munari 2000).}
181: \label{Example_GalQSO_Spec}
182: \end{figure}
183: %
184: This plot clearly indicates that the lensing galaxy can modify
185: the near-IR to visible light colour of lensed quasar systems.
186: In this paper, we adopt the $g-H$ colour as a measure of the near-IR excess.
187:
188: In order to exploit this idea to look for new lensed quasars in the
189: SDSS, we cross-correlated the SDSS Data Release 5 (DR5) spectroscopic
190: quasars\footnote{In this paper we do not use the DR5 quasar catalog of
191: Schneider et al. (2007), because this work began before
192: the public release of their catalog.} with the Two Micron All Sky Survey
193: (2MASS) point sources (Cutri et al. 2003). Specifically we looked for a
194: 2MASS point source within $5''$ from each quasar position and extract
195: its IR magnitudes\footnote{We make this catalog available to the
196: community through the Vizier service:
197: http://vizier.u-strasbg.fr/cgi-bin/VizieR. The online catalog includes
198: also cross correlation with the FIRST radio catalog (Becker et
199: al. 1995).}.
200: %
201: For our lens search, we have constructed a subsample of candidates by
202: selecting quasars with: redshift $3>z>1$; redshift confidence
203: z\_conf $>0.9$; $i$-band magnitude brighter than 19.0; and angular
204: distance between the SDSS and 2MASS sources smaller than
205: $1\farcs5$. We excluded low-redshift quasars because their fluxes
206: are sometimes affected to a large extent by contributions
207: from host galaxies. We also avoided selecting high redshift quasars
208: since the lens galaxies of such quasars will be typically at high
209: redshifts, and therefore are too faint to be detected in the 2MASS
210: survey (see \S\ref{MethodNearIR}).
211: %
212: %
213: The smaller
214: angular distance cutoff (i.e., $1\farcs5$)
215: selects $95\%$ of all the
216: SDSS-2MASS sources found up to $5''$ apart. We use the relatively small
217: offset in order to remove well separated quasar-galaxy pairs. The
218: resulting number of SDSS-2MASS quasars is 3132.
219: %
220: \begin{figure}
221: \centerline{\epsfxsize=8.5cm\epsfbox{f2.eps}}
222: \caption{The optical-IR colour $g-H$ versus redshift for 3347 ($=3132+215$)
223: SDSS-2MASS quasars (dots) with SDSS measured redshifts above 1,
224: redshift confidence $>90\%$, $i<19.0$~mag, and angular distance between the SDSS and
225: 2MASS sources smaller than $1\farcs5$. Also shown are the positions of known
226: lensed quasars (filled circles) and of SDSS~J$131339.98+515128.3$ (pentagram).}
227: \label{SDSS2MASS_gH_z}
228: \end{figure}
229: %
230: \begin{figure}
231: \centerline{\epsfxsize=8.5cm\epsfbox{f3.eps}}
232: \caption{Same as Figure~\ref{SDSS2MASS_gH_z}, but for $g-H$ versus
233: $g-r$. The line marks our selection cut in this colour space (see text).}
234: \label{SDSS2MASS_gH_gr}
235: \end{figure}
236: %
237: In Figures~\ref{SDSS2MASS_gH_z} and~\ref{SDSS2MASS_gH_gr}
238: we show the locations (dots) of the 3132 SDSS-2MASS quasars passing these
239: selection criteria, in addition to 215
240: quasars found above redshift 3,
241: in the redshift versus $g-H$ colour and $g-r$
242: versus $g-H$ colour phase spaces, respectively.
243: Also shown are the positions of known lensed quasars with SDSS and
244: 2MASS detections (filled circles), and the position
245: of the new lens reported in this paper (SDSS~J131339.98$+$515128.3;
246: pentagram).
247: %
248: Figure~\ref{SDSS2MASS_gH_gr} indicates that the optical-IR colour $g-H$
249: is correlated with the optical colour $g-r$, reflecting the fact that the
250: quasar spectral energy distributions are characterized by a power-law.
251: %different indices for different quasars (REF).
252: The Figure suggests
253: that known lenses tend to be redder than the ``average'' quasar.
254: %If some of these $g-H$ outliers are caused by the contribution
255: %of a lensing galaxy, this can be used to select lensed quasars,
256: %especially in unresolved systems.
257:
258: In order to test the feasibility of this method, from the 3132
259: quasars, we selected objects whose colours satisfy:
260: $g-H>1.15(g-r)+2.3$ (above the line in
261: Figure~\ref{SDSS2MASS_gH_gr}),
262: where, $g$ and $r$ are the $g$ and $r$-band
263: magnitudes corrected for Galactic extinction (Schlegel et al. 1998;
264: Cardelli et al. 1989).
265: This colour cutoff is arbitrary and does not rely on any
266: quantitative argument.
267: We choose this cut as an initial trial to study
268: the feasibility and efficiency of the near-IR-excess lensed quasar selection method.
269: %At this point we restrict quasars with
270: %redshift $z<3$, because very high redshift quasars have quite different
271: %$g-r$ colour with lower redshift quasars.
272: Using the $g-H$ cut we selected
273: 294 quasars, of which the following 7 are previously known lenses:
274: HS0810+2554 (Reimers et al. 2002);
275: SBS0909+523 (Kochanek et al. 1997; Oscoz et al. 1997);
276: SDSS~J1155+6346 (Pindor et al. 2004);
277: SDSS~J1206+4332 (Oguri et al. 2005);
278: SDSS~J1226$-$0006 (Inada et al. in preparation);
279: SDSS~J1335+0118 (Oguri et al. 2004);
280: SDSS~J1524+4409 (Oguri et al., in preparation).
281: Therefore the lens search efficiency is already $>2\%$ at
282: this point, before introducing any morphological constraints.
283:
284: We make the lens survey more efficient,
285: and reduce the number of targets for follow up observations,
286: by applying a simple cut on the
287: morphology of the quasars, in the same spirit as Oguri et al. (2006).
288: Specifically, we set the condition that the SDSS stellar likelihood
289: of the candidates is
290: smaller than $0.1$ in both the $r$ and $i$-bands. The low stellar
291: likelihood of $<0.1$ means that the object is poorly fitted by the point
292: spread function (PSF).
293: %
294: We note that the completeness for including extended quasars
295: in the spectroscopic quasar sample begins to
296: decrease at $z=2.2$, and at $z\gtorder3$ most extended
297: quasars are not targeted (Oguri et al. 2006).
298: %
299: %We introduce the stellar likelihood
300: %criteria in order to reduce the number of targets for follow-up.
301: We finally compiled a list of 56 candidates
302: that survived the morphology cut (14 of them with $z>2.2$).
303: We note that all the 7 previously known lenses pass the colour and morphology
304: criteria.
305: Therefore, the lens selection efficiency improves to more than $10\%$.
306:
307: Next, we look for new gravitationally
308: lensed quasars in this list of candidates.
309: We exclude seven of the radio-loud targets in
310: the list, because they were observed at $0\farcs22$ resolution
311: in the CLASS radio lens survey
312: and confirmed not to be lenses (e.g., Myers et al. 2003).
313: We observed two targets at the Subaru 8.2\,m telescope
314: and 12 targets at the University of Hawaii 2.2\,m telescope, and
315: found that one of the targets, SDSS~J$131339.98+515128.3$, consists
316: of two optical point sources and an extended source, making it an excellent lens
317: candidate. SDSS~J$131339.98+515128.3$ has no detectable radio flux at
318: the $\sim1$~mJy limit of the FIRST 20-cm survey (Becker et al. 1995).
319:
320: %We inspected the SDSS frames of these quasars, and selected
321: %the best candidates which show indications for duplicity.
322:
323:
324:
325:
326: \section{Observations of SDSS~J131339.98$+$515128.3}
327: \label{Obs}
328:
329: On UTC 2007 April 12.95, we observed our lens candidate
330: SDSS~J$131339.98+515128.3$ with
331: the Tektronix $2048\times2048$ CCD camera (Tek2k)
332: at the University of Hawaii $2.2$\,m telescope, in $VRI$-bands. The
333: exposure time was 300~s for each band.
334: %
335: The frames with a pixel scale of $0\farcs 2195$~pix$^{-1}$,
336: were obtained under seeing of about $0\farcs8$, and
337: are shown in Figure~\ref{SDSS1313_zoom}.
338: The data were reduced using standard IRAF\footnote{The Image Reduction
339: and Analysis Facility (IRAF), is general purpose
340: software system for the reduction and analysis of astronomical
341: data. IRAF was written by the programming group at
342: the National Optical Astronomy Observatories (NOAO) in Tucson,
343: Arizona. NOAO is operated by the Association of Universities for
344: Research in Astronomy (AURA), Inc. under cooperative agreement
345: with the National Science Foundation} tasks, and
346: photometric calibration was preformed using observations of the
347: Landolt (1992) standard star fields PG0918$+$029 and PG1633$+$099.
348: %
349: %\begin{figure}
350: %\centerline{\epsfxsize=8.5cm\epsfbox{f4.eps}}
351: %\caption{$I$-band, 300~s, finding chart of the field of
352: %SDSS~J$131339.98+515128.3$, obtained with the Tek2k at the
353: %University of Hawaii $2.2$~m telescope.}
354: %\label{SDSS1313_I_FC}
355: %\end{figure}
356: %
357:
358:
359: \begin{figure}
360: \centerline{\epsfxsize=8.5cm\epsfbox{f5.eps}}
361: \caption{{\bf Upper panels:} $VRI$ frames (from left to right)
362: around SDSS~J$131339.98+515128.3$, obtained at
363: the University of Hawaii $2.2$~m telescope. The frames are presented
364: with the same gray-scale stretch, and the colour difference between the
365: two quasar images is apparent. Throughout this paper
366: the brighter quasar image (south) is denoted A and the fainter
367: (north) is denoted B.
368: {\bf Lower panels:} $VRI$ frames (from
369: left to right) of the lensed system after subtracting the two quasar
370: images. The lensing galaxy is clearly detected. The frames are displayed
371: with the same gray-scale stretch, which is about $1/6$ of the
372: gray-scale stretch in the upper panels. The white dots are residuals
373: due to the subtraction of the lensed quasar images.}
374: \label{SDSS1313_zoom}
375: \end{figure}
376: %
377:
378:
379: On UTC 2007 March 10.56 we obtained a 1400\,s spectrum of
380: SDSS~J$131339.98+515128.3$ using the Low Resolution Imaging
381: Spectrometer (LRIS; Oke et al. 1995) on the Keck-I 10\,m telescope.
382: In the red arm we used the 400/8500 grating, and on the blue side we
383: used the 400/3400 grism. The spectrum was obtained using the $0\farcs7$
384: slit, at a position angle of $16$\,deg, which goes through the two
385: lensed quasar images. For flux calibration we observed the
386: spectrophotometric standard star HZ~44. The spectra were reduced using
387: tools developed in the MATLAB environment (Ofek et al. 2006).
388:
389: \section{Photometry and Spectra}
390: \label{Phot}
391:
392: The positions and magnitudes of the lensed quasar images and lensing
393: galaxy are derived by fitting the system with two PSFs and a galaxy
394: profile using Galfit (Peng et al. 2002). We first model the galaxy by a
395: Sersic profile and found the best-fit Sersic index to be $n\approx 4$.
396: Therefore, we fix the index to $n=4$, a typical value for early type
397: galaxies, to compute the relative positions and magnitudes.
398: % However, we note that the galaxy is faint and therefore,
399: % can be fit relatively well with other profiles (e.g., exponential profile).
400: Tables~\ref{Tab-Pos} and \ref{Tab-Mag} summarize the fit results.
401: %
402: The separation of the lensed quasar images is $1\farcs24$, and the
403: flux ratios between the two quasar images are $3.63$,
404: $1.67$ and $0.92$ in the $VRI$-bands, respectively.
405: %
406: \begin{table}
407: \begin{center}
408: \begin{tabular}{lcc}
409: \hline
410: Object & $\Delta\alpha$& $\Delta\delta$ \\
411: & [arcsec.] & [arcsec.] \\
412: \hline
413: A & $ 0.0 $ & $ 0.0 $\\
414: B & $-0.325 \pm 0.002$ & $ 1.193 \pm 0.002$\\
415: G & $-0.064 \pm 0.009$ & $ 0.153 \pm 0.009$\\
416: \hline
417: \end{tabular}
418: \caption{The relative positions were calculated from the mean
419: of the $VRI$ image positions and the errors were estimated from the scatter.
420: The absolute position of image A is: 13:13:39.98 $+$51:52:28.37 (J2000.0).}
421: \label{Tab-Pos}
422: \end{center}
423: \end{table}
424: %
425: \begin{table}
426: \begin{tabular}{lcccccc}
427: \hline
428: Object & $V$ & $R$ & $I$ & $b/a$ & PA & $R_{e}$ \\
429: & [mag] & [mag] & [mag] & & [deg] & [arcsec.] \\
430: \hline
431: A & $18.26$ & $18.09$ & $17.45$ & & & \\
432: B & $19.66$ & $18.65$ & $17.36$ & & & \\
433: G & $18.74$ & $18.23$ & $17.49$ & $0.78$ & $-71$ & $0.75$ \\
434: \hline
435: \end{tabular}
436: \caption{The $VRI$ magnitudes of the lensed quasar images and the
437: lensing galaxy as obtained by two PSFs$+$galaxy fitting. The magnitudes are
438: given in the Vega system. We fitted a Sersic profile to the galaxy
439: using Galfit (Peng et al. 2002). The best fit Sersic index is $\sim4$.
440: $R_{e}$ is the effective radius of the galaxy.}
441: \label{Tab-Mag}
442: \end{table}
443: %
444: We note that the lensing galaxy is also apparent in an $R$-band
445: image taken using the Suprime-cam on the Subaru telescope (Miyazaki et
446: al. 2002), which was retrieved from the Subaru telescope archive (Baba
447: et al. 2002). However we did not use the Subaru data for our analysis
448: because the PSF of the image is very unusual, which prevents reliable
449: photometry and astrometry of the system.
450:
451: The spectra of the two images of SDSS~J$131339.98+515128.3$
452: are displayed in Fig~\ref{SDSS1313_SpecMain},
453: and the flux ratio, as a function of wavelength,
454: between the spectra of images A and B is
455: presented in Fig.~\ref{SDSS1313_SpecRatio}.
456: %
457: \begin{figure}
458: \centerline{\epsfxsize=8.5cm\epsfbox{f6.eps}}
459: \caption{The spectra of SDSS~J$131339.98+515128.3$
460: images A (upper spectrum) and B (lower spectrum).
461: The three insets show absorption lines of
462: Mg~II ($\lambda\lambda 2798.1$-$2802.7$~\AA\AA),
463: Ca~II ($\lambda\lambda 3934.8$-$3969.6$~\AA\AA), and
464: Na~I ($\lambda\lambda 5889.9$-$5895.9$~\AA\AA),
465: in the spectra, presumably due to the lensing galaxy (see
466: Figure~\ref{SDSS1313_zoom}). The gray boxes mark the position of the
467: insets in the spectrum. The gaps in the spectra are due to removal of
468: telluric absorption lines.}
469: \label{SDSS1313_SpecMain}
470: \end{figure}
471: %
472: \begin{figure}
473: \centerline{\epsfxsize=8.5cm\epsfbox{f7.eps}}
474: \caption{The ratio between the spectra of SDSS~J$131339.98+515128.3$
475: images A and B. The positions of the quasar broad emission lines
476: are marked by arrows.}
477: \label{SDSS1313_SpecRatio}
478: \end{figure}
479: %
480: The two quasars, marked by A and B, have identical redshifts with
481: similar (though not identical)
482: broad emission lines, but markedly different continua.
483: We will discuss the possible
484: origins of this large colour difference in \S\ref{Disc}.
485: %
486: In the spectra we detect three absorption doublets of Mg~II,
487: Ca~II and Na~I (insets in Figure~\ref{SDSS1313_SpecMain})
488: at a redshift of $0.194\pm0.005$, which are probably induced by
489: the lensing galaxy. Additional support for this interpretation comes
490: from the fact that the stellar absorption features of Ca H\&K are
491: stronger for image A, which is closer to the lensing galaxy.
492:
493: We fitted the Galactic extinction (Schlegel et al. 1998)
494: corrected colours of the galaxy with synthetic photometry
495: of galaxy spectral templates (Kinney et al. 1996) assuming the lens
496: redshift of $0.194$. The visible-light colours of the lensing galaxy
497: are best fit with a late-type galaxy.
498: However, these colours may be affected by the light of
499: image A, which is only $0\farcs17$ from the
500: position of the galaxy.
501: The late-type galaxy colours may contradict the
502: best-fit Sersic index of $\sim 4$.
503: %
504: To explore this issue, we tried fitting
505: the lens galaxy with an $n=1$ Sersic profile, and found
506: that the fit to the imaging data is reasonable.
507: Therefore, further observations are needed for a more conclusive
508: classification of the lensing galaxy.
509:
510: \section{Lens modeling}
511: \label{Model}
512:
513: We fitted, in the source plane, the image positions with an elliptical
514: softened power law mass model (Barkana 1998). In the fit we assumed a
515: core radius of 1 milliarcsecond (mas)
516: and used the following priors:
517: the lens position to lie within $27$~mas from the galaxy center (about $3$
518: times the errors in the galaxy position);
519: the lens eccentricity to be less than $0.35$;
520: the position angle (PA) to lie between $-40$ and $-100$ deg
521: (which is $\sim30$~deg from the PA of the galaxy light distribution);
522: and the power-law of the two-dimensional mass
523: distribution, $\alpha$, to be $1.0$ (corresponding to an isothermal sphere).
524: %
525: The best fit parameters are: centre-of-mass position
526: $\Delta$R.A.$=-0\farcs048$, $\Delta$Dec.$=0\farcs175$ (relative to image A);
527: eccentricity $0.16$; PA$=-70$~deg; and the position of
528: the source is $\Delta$R.A.$=-0\farcs163$, $\Delta$Dec.$=0\farcs595$.
529: %
530: The time delay, estimated using the analytic formula for an isothermal
531: potential (Kassiola \& Kovner 1993), is around 30 days, assuming the third year
532: WMAP cosmology\footnote{$H_{0}=70.4$~km~s$^{-1}$~Mpc$^{-1}$; $\Omega_{m}=0.27$;
533: $\Omega_{\Lambda}=0.73$.} (Spergel et al. 2007).
534:
535: Given these parameters, the derived magnifications are
536: $-0.4$ and $2.5$ for images A and B, respectively.
537: These predicted macro-model magnifications are inconsistent with the
538: observed flux ratio (i.e., the nearest image to the lens is the
539: brightest), and this discrepancy is discussed in the next section.
540:
541:
542:
543:
544: \section{Discussion}
545: \label{Disc}
546:
547: Here we discuss the efficiency of our near-IR excess method
548: (\S\ref{EffNearIR}), its selection effects (\S\ref{MethodNearIR}),
549: as well as the nature of the different colours of the lensed images of
550: SDSS~J$131339.98+515128.3$ (\S\ref{DiscColor}).
551:
552:
553: \subsection{Efficiency of the near-IR excess method}
554: \label{EffNearIR}
555:
556: Our preliminary selection criteria in the $g-r$ versus $g-H$ colour space
557: selected 294 candidates, of which at least 8 are gravitational lenses.
558: The efficiency $\ge8/294\sim 3\%$ is apparently higher than lensing rates in
559: quasar samples. For comparison, the lensing efficiency of the CLASS lens survey
560: was $\sim0.14\%$ (Browne et al. 2003), and about $1\%$ in
561: the Hubble Space Telescope snapshot survey (e.g., Maoz et al. 1993).
562: Therefore we conclude that our near-IR
563: excess cut indeed enhances the fraction of lenses in the sample.
564: %We note
565: %that the sample was constructed solely by the colour cuta and no condition on
566: %the morphology of quasars was placed.
567: After the introduction of the morphology selection criterion,
568: the efficiency boosts to $\ge8/56\sim 14\%$. However,
569: such a morphology cut may drop very small separation lensed quasars that
570: are barely resolved in the SDSS.
571:
572:
573: We expect the efficiency of this method to increase for intrinsically
574: fainter quasars in which the lensing galaxy has a higher probability of
575: dominating the system light.
576: %However lens surveys of such faint quasars
577: %with the near-IR method will require deep wide-field IR surveys.
578: %
579: We comment that a significant fraction
580: of false positive candidates
581: in our near-IR excess sample are close
582: pairs of quasars and galaxies which are inevitably selected given the
583: philosophy of the method.
584:
585:
586: \subsection{Selection effects of the near-IR excess method}
587: \label{MethodNearIR}
588:
589: The near-IR excess selection method is biased toward systems in which
590: the lensing galaxy is apparently bright, and therefore, on average at
591: lower redshift than the typical redshift of lensing galaxies.
592: %
593: Moreover, these systems
594: will have larger than average image separations.
595: This is a consequence of the fact that
596: a brighter, and hence more massive, lensing galaxy implies
597: a larger image separation.
598: The lower lens redshift also
599: results in larger image separation.
600: %
601: %Indeed the lensing galaxy
602: %of our new lensed quasar is located at relatively low-redshift
603: %($z=0.194$). Another example of this is SDSS J1524+4409 whose lensing
604: %galaxy is dominant even at visible-light wavelengths.
605: Therefore, caution is needed when using lens systems found by such a
606: method in analysis which requires lensed quasar samples with
607: non-biased lens redshift distribution (e.g., Kochanek 1992; Ofek et
608: al. 2003; Capelo \& Natarajan 2007).
609:
610:
611: In addition, this method will probably be biased toward detecting doubly
612: imaged lensed quasars in which the average magnification is smaller than
613: in quadruply lensed systems, and therefore, the lensing galaxy has
614: a higher probability to dominate the total light in the near-IR.
615: %
616: We note that in a red-band flux-limited sample, like
617: the one we use here (i.e., $i<19$~mag), the incidence of lenses
618: will be increased due to the contribution of the lens
619: galaxy to the total system light,
620: relative to a sample which is not flux limited.
621:
622:
623: The images of some lensed quasars show remarkable rest-frame UV
624: continuum differences (e.g., HE1104$-$1805 Wisotzki et al. 1993) that
625: are variable with time (e.g., Ofek \& Maoz 2003; Poindexter et al. 2007).
626: Such continuum differences can make them appear redder.
627: %
628: Although this type of lensed quasars is relatively rare, it is
629: possible that the new quasar presented in this paper belongs to this group.
630: %
631: Furthermore, it is well known that narrow and broad absorption line QSOs
632: are redder than typical quasars (e.g., Brotherton et al. 2001), and this
633: is especially true for radio selected quasars (e.g., Menou et al. 2001).
634: Therefore, the near-IR excess method may be
635: efficient for looking for lensed broad absorption line quasars, which
636: may be useful for probing the structure of quasar outflows using
637: microlensing (Chelouche 2005).
638:
639:
640: Gregg et al. (2002) argued that the population of very red quasars
641: contains a large fraction of gravitational lenses because of reddening
642: by the lensing galaxies.
643: %
644: Although we compare optical-IR colours ($g-H$) of quasars with the optical
645: ($g-r$) colours in order to distinguish between red/reddened quasars and
646: lensed quasar systems, it is still possible that this method is biased
647: toward selecting red lensed quasars. This is also suggested by the fact
648: that one of the lensed components of our new lens has a red colour.
649: %
650: Note, however, that highly extincted lenses are
651: relatively rare (e.g., Falco et al. 1999).
652:
653:
654: %
655: Finally, we note that a basic limitation of this approach is that it can
656: detect the lensing galaxy only if it is brighter than the limiting
657: magnitude of the near-IR survey that is being used. In particular it
658: should be emphasized that the current use of this method is limited by
659: the shallow detection limit of the 2MASS survey used here.
660: In the 2MASS survey $5L_{*}$ galaxies can be detected up to a redshift
661: of $\sim0.2$.
662: Thus, lensed systems with higher-redshift lensing galaxies
663: cannot be efficiently detected using this method. However,
664: future surveys
665: (e.g., UKIDSS, Lawrence et al. 2006; WISE, Mainzer et al. 2005)
666: can be used to
667: select lensing galaxies to redshifts up to $\sim 1.0$, making the
668: near-IR excess method much more useful.
669: %
670: % L_* 0.1 0.5
671: % 5*L* to z=0.2 and 0.95 using 2mass and UKIDSS...
672: %
673: %Finally, we note that the lensed quasars
674: %FIRST~J100424.9$+$122922 (Lacy et al. 2002) was
675: %discovered based on its near IR excess, as part
676: %of a survey for red quasars (Gregg et al. 2002).
677:
678:
679: \subsection{The origin of the different colours of the lensed images}
680: \label{DiscColor}
681:
682: The lensed images of SDSS~J$131339.98+515128.3$ are distinctly
683: different in their spectral continuum slopes, but the shapes and
684: equivalent widths of their broad emission lines are similar (though not
685: identical).
686: %
687: Such colour differences are not common among lensed quasars,
688: although there are several known examples (e.g., HE1104$-$1805).
689: %
690: Two obvious possibilities are that the different colours
691: % (see Table~\ref{Tab-Mag})
692: are caused by microlensing, or/and by
693: extinction and reddening in the
694: lensing galaxy (e.g., Falco et al. 1999; El{\'{\i}}asd{\'o}ttir et
695: al. 2006).
696:
697: %It is well known that the equivalent width of the
698: %Na~I absorption line doublet is correlated with the extinction
699: %(e.g., Turatto et al. 2003).
700: The equivalent widths of the Na~I absorption lines in the spectra
701: of images A and B (Figure~\ref{SDSS1313_SpecMain}), are
702: $\sim2.5$~\AA~and $\sim1.5$~\AA, respectively.
703: %
704: The known correlation between Na~I absorption lines
705: and extinction (e.g., Turatto et al. 2003)
706: suggests that the lensed quasar images
707: are extincted
708: with $E_{B-V}$ in the range of $0.1$ to $1$~mag.
709:
710:
711:
712:
713: To estimate the relative extinction between
714: the two quasar images, we divided the continuum-subtracted
715: spectrum of image A by that of image B
716: (i.e., we used only the broad line flux),
717: and fit it with extinction curves (Cardelli et al. 1989).
718: %
719: %For estimating the extinction we used only the flux
720: %in the broad emission lines, since it is
721: %less susceptible to microlensing (see below).
722: %
723: The best fit extinction we have found is $E_{B-V}=0.5$~mag
724: (assuming $R_{V}=3.08$).
725: %
726: Next, we divided the spectrum of image A by
727: the de-reddened spectrum of image B (i.e., $E_{B-V}=0.5$~mag, $R_{V}=3.08$).
728: The result is roughly consistent with a power-law,
729: $\sim\lambda^{-1\pm0.1}$.
730: This indicates that extinction alone probably cannot
731: explain the entire difference between the spectra.
732: %
733: We note that if we use the entire spectra (continuum and lines)
734: to fit for a relative extinction curve between
735: the two lensed images, a reasonable fit is found
736: with $E_{B-V}\approx0.7$~mag.
737: However, this does not constitute a perfect fit,
738: and the residuals between the spectra
739: are not monotonic with wavelength, which is difficult to explain.
740:
741:
742:
743: Contrary to extinction,
744: we expect microlensing
745: to alter the spectral continuum of a
746: quasar image, but keep the broad emission
747: lines unchanged.
748: %
749: The reason for this is that
750: the approximate length scale over which microlensing effects are
751: expected to average out, is several times the microlensing Einstein
752: radius. The expected Einstein radius for 1~M$_{\odot}$
753: microlensing in this system is about $0.03$~pc (in the source plane),
754: which is of the same order of magnitude as the expected size
755: of the accretion disks of bright quasars.
756:
757:
758: Interestingly, as shown in
759: Fig~\ref{SDSS1313_SpecRatio},
760: the ratio between the lensed images' broad emission lines
761: is not identical to the ratio of the continua around these lines.
762: This can be investigated
763: further by a detailed comparison of the line profiles,
764: shown in Fig.~\ref{SDSS1313_EmissionLinesComparison}.
765: Here we have made the assumption that the
766: relative reddening between
767: images A and B is $E_{B-V}=0.5$~mag.
768: Then, for each broad emission line,
769: we have multiplied the de-reddened spectrum of image B
770: by a wavelength dependent factor,
771: %
772: %by 1.5, 1.0, 0.93 and 0.64,
773: %for the Ly$\alpha$, C~IV, C~III], and Mg~II, respectively.
774: %This was done
775: %
776: in order to match the continua of images
777: A and B around these lines.
778: %
779: As shown in Fig.~\ref{SDSS1313_EmissionLinesComparison},
780: we find remarkably good agreement between
781: the lines of images A (black) and B (gray),
782: in the case of the C~III] and Mg~II lines.
783: In the other two lines (Ly$\alpha$ and C IV) the
784: broad wings of the line also agree, but there is clearly a narrower
785: component present within the broad emission lines of
786: image B that is absent in image A.
787:
788:
789: There are two possible explanations for this result. The first is that
790: the broader component of the emission lines
791: originates in a region small enough to be
792: microlensed, together with the optical continuum, in the sense that both
793: continuum and broad component of the emission lines in image
794: B are suppressed.
795: %
796: We note that microlensing of a quasar's broad emission lines
797: was probably detected in SDSS~1004$+$4112
798: (Richards et al. 2004).
799: %
800: The second possibility is that reddening is mainly responsible for this
801: difference.
802: In this case we would
803: require that the continuum and broad line components
804: are affected by a different degree of reddening
805: than the narrow component of the emission lines.
806: In order for extinction in the lensing galaxy to produce a differential
807: effect of this type, it should be
808: patchy on a length scale of the broad-line region size (i.e., $\sim1$~pc).
809: %
810: %Moreover, the redder image (B) is found further away
811: %from the galaxy ($1\farcs06$) than image A ($0\farcs35$),
812: %and the equivalent width of the Na~I absorption lines
813: %is marginally stronger in image A.
814: %
815: %Therefore,
816: %
817: \begin{figure}
818: \centerline{\epsfxsize=8.5cm\epsfbox{f9.eps}}
819: \caption{Each panel shows one of the major emission lines in the
820: spectrum: Ly$\alpha$, C~IV, C~III] and Mg~II, at the observed
821: wavelengths.
822: In each case,
823: the black-line shows the spectrum of image A,
824: and the gray line represents the observed,
825: extinction corrected ($E_{B-V}=0.5$, $R_{V}=3.08$),
826: spectrum of image B, multiplied
827: by 1.5, 1.0, 0.93 and 0.64,
828: for the Ly$\alpha$, C~IV, C~III], and Mg~II, respectively.
829: It can be seen that this scaling gives a good overlap for Mg~II and
830: C~III] lines, and for the broad components of the other lines, but that there
831: are narrow components in Ly$\alpha$ and C~IV lines,
832: which do not overlap well,
833: being considerably greater in image B.}
834: \label{SDSS1313_EmissionLinesComparison}
835: \end{figure}
836:
837:
838:
839:
840: The most simple explanation for the spectral differences
841: between the lensed quasar images, is that while they
842: are affected by extinction,
843: microlensing plays a significant role.
844: %
845: The spectral differences, after applying a
846: relative extinction of $E_{B-V}=0.5$~mag between the images,
847: suggests that the quasar region
848: responsible for the short-wavelength
849: ($\lambda_{rest}\sim1400$~\AA)~continuum emission
850: is more magnified than the region emitting
851: the longer-wavelength
852: ($\lambda_{rest}\sim2800$~\AA)~radiation.
853: %
854: Indeed, this is consistent with what is expected from
855: microlensing of a quasar accretion disk.
856: %
857: Moreover,
858: %
859: %We note that
860: %Mortonson et al. (2005)
861: %calculated the microlensing magnification probability
862: %distribution of a quasar accretion disk for the negative and positive
863: %parity images.
864: %
865: %
866: the negative parity image (image A in our case) has a higher probability
867: of being magnified by microlensing relative to the positive parity
868: image (image B; Schechter \& Wambsganss 2002; Mortonson et al. 2005).
869: Indeed, contrary
870: to the macro-model (\S\ref{Model}) prediction, image A is magnified
871: relative to image B.
872: %
873: Probably the best way to verify that
874: microlensing is responsible for some of the
875: spectral differences between the lensed images, is to look
876: for temporal colour changes between the two lensed images,
877: similar to the colour variability observed
878: in HE1104$-$1805 (Poindexter et al. 2007).
879: %
880: %It will also be important to use data at longer wavelengths to assess
881: %the intrinsic flux ratio of the two lensed images.
882: %
883: %We conclude that the lensed quasar images
884: %spectra are modified by both extinction and microlensing.
885: %Final differentiation between the effects of
886: %microlensing and extinction must await further studies.
887:
888:
889:
890:
891:
892:
893:
894: \section{Conclusion}
895: \label{Conc}
896:
897: Given the promising future of whole sky surveys (e.g., SkyMapper, Keller
898: et al. 2007; PanSTTAR, Kaiser et al. 2005; LSST, Tyson 2005), a variety
899: of methods to search for gravitationally lensed quasars are needed.
900: %
901: One possibility, suggested by Kochanek et al. (2006), is to look for
902: variable sources that are extended, presumably due to lensed quasar
903: images and/or lensing galaxy. However, this method, like the SDSS quasar
904: lens search (Oguri et al. 2006), is limited by the seeing of ground based
905: telescopes.
906:
907: We have proposed that some classes of lensed quasars can be identified
908: efficiently by looking for single-epoch near-IR excess of known
909: (spectroscopic or photometric) quasars. The near-IR method is based on
910: the fact that at least some lensed quasar systems will be redder due to
911: the presence of a bright lensing galaxy (e.g., Fig~\ref{Example_GalQSO_Spec}).
912: %
913: The use of the near-IR excess method is demonstrated by our discovery of
914: the new lensed quasar SDSS~J$131339.98+515128.3$. Follow up observations
915: have shown it to be a doubly-imaged lensed quasar, at $z=1.875$, with an
916: image separation of $1\farcs24$. We also identified the lensing galaxy
917: at a redshift of $0.194$. The continua of the lensed quasar images are
918: markedly different, presumably due to extinction and microlensing effects.
919: %because of microlensing of the accretion disk of one of the quasar
920: %images or/and differential reddening.
921: The discovery of the new lens as well as the presence of many known
922: lensed quasars in the near-IR excess lens candidates sample suggests
923: that the near-IR excess is indeed an efficient way to select lensed
924: quasar systems from large samples of quasars.
925:
926: \section*{ACKNOWLEDGMENTS}
927: EOO, MO, and NJ would like to thank the Kavli Institute for
928: Theoretical Physics and the organizers of the KITP workshop
929: ``Applications of Gravitational Lensing: Unique insights into Galaxy
930: Formation and Evolution'' for their hospitality; this work was initiated
931: at this KITP workshop.
932: It is pleasure to thank the referee, Chris Kochanek, for useful
933: comments on the manuscript.
934: This research was supported in part by the
935: National Science Foundation under Grant No. PHY05-51164, and also by
936: Department of Energy contract DE-AC02-76SF00515 and by the European
937: Community's Sixth Framework Marie Curie Research Training Network
938: Programme contract no. MRTN-CT-2004-505183 ``ANGLES''. Use of the UH 2.2-m
939: telescope for the observations is supported by NAOJ.
940: Based in part on data collected at Subaru telescope and obtained
941: from the SMOKA archive, which is operated by
942: the Astronomy Data Center, National Astronomical Observatory of Japan.
943: NI acknowledges support from the Special Postdoctoral
944: Researcher Program of RIKEN.
945:
946:
947:
948:
949:
950: \begin{thebibliography}{}
951:
952: %\bibitem[Alard \& Lupton(1998)]{1998ApJ...503..325A} Alard, C., \& Lupton,
953: %R.~H.\ 1998, ApJ, 503, 325
954:
955: \bibitem[Baba et al.(2002)]{2002RNAOJ...6...23B} Baba, H., et al.\ 2002,
956: Report of the National Astronomical Observatory of Japan, 6, 23
957:
958: \bibitem[Barkana(1998)]{1998ApJ...502..531B} Barkana, R.\ 1998, ApJ, 502,
959: 531
960:
961: \bibitem[Becker et al.(1995)]{1995ApJ...450..559B} Becker, R.~H., White,
962: R.~L., \& Helfand, D.~J.\ 1995, ApJ, 450, 559
963:
964: \bibitem[Borgeest et al.(1991)]{1991A&A...251L..35B} Borgeest, U., von
965: Linde, J., \& Refsdal, S.\ 1991, A\&A, 251, L35
966:
967: \bibitem[Brotherton et al.(2001)]{2001ApJ...546..775B} Brotherton, M.~S.,
968: Tran, H.~D., Becker, R.~H., Gregg, M.~D., Laurent-Muehleisen, S.~A., \&
969: White, R.~L.\ 2001, ApJ, 546, 775
970:
971: \bibitem[Browne et al.(2003)]{2003MNRAS.341...13B} Browne, I.~W.~A., et
972: al.\ 2003, MNRAS, 341, 13
973:
974: \bibitem[Capelo \& Natarajan(2007)]{2007arXiv0705.3042C} Capelo, P.~R., \&
975: Natarajan, P.\ 2007, astro-ph/0705.3042
976:
977: \bibitem[Cardelli et al.(1989)]{1989ApJ...345..245C} Cardelli, J.~A.,
978: Clayton, G.~C., \& Mathis, J.~S.\ 1989, ApJ, 345, 245
979:
980: \bibitem[Chelouche (2005)]{} Chelouche D., 2005, ApJ, 629, 667
981:
982: \bibitem[Cohn et al.(2001)]{2001ApJ...554.1216C} Cohn, J.~D., Kochanek,
983: C.~S., McLeod, B.~A., \& Keeton, C.~R.\ 2001, ApJ, 554, 1216
984:
985: \bibitem[Cutri et al.(2003)]{2003tmc..book.....C} Cutri, R.~M., et al.\
986: 2003, The IRSA 2MASS All-Sky Point Source Catalog, NASA/IPAC Infrared
987: Science Archive.~http://irsa.ipac.caltech.edu/applications/Gator/,
988:
989: \bibitem[El{\'{\i}}asd{\'o}ttir et al.(2006)]{2006ApJS..166..443E}
990: El{\'{\i}}asd{\'o}ttir, {\'A}., Hjorth, J., Toft, S., Burud, I., \&
991: Paraficz, D.\ 2006, ApJS, 166, 443
992:
993: \bibitem[Falco et al.(1999)]{1999ApJ...523..617F} Falco, E.~E., et al.\
994: 1999, ApJ, 523, 617
995:
996: \bibitem[Gregg et al.(2002)]{2002ApJ...564..133G} Gregg, M.~D., Lacy, M.,
997: White, R.~L., Glikman, E., Helfand, D., Becker, R.~H., \& Brotherton,
998: M.~S.\ 2002, ApJ, 564, 133
999:
1000: \bibitem[Jackson \& Browne(2007)]{2007MNRAS.374..168J} Jackson, N., \&
1001: Browne, I.~W.~A.\ 2007, MNRAS, 374, 168
1002:
1003: \bibitem[Johnston et al.(2003)]{johnston03} Johnston, D.~E., et
1004: al.\ 2003, AJ, 126, 2281
1005:
1006: \bibitem[Kaiser \& Pan-STARRS Collaboration(2005)]{2005AAS...206.2402K}
1007: Kaiser, N., \& Pan-STARRS Collaboration 2005, Bulletin of the American
1008: Astronomical Society, 37, 465
1009:
1010: \bibitem[Kassiola \& Kovner(1993)]{1993ApJ...417..450K} Kassiola, A., \&
1011: Kovner, I.\ 1993, ApJ, 417, 450
1012:
1013: \bibitem[Keller et al.(2007)]{2007PASA...24....1K} Keller, S.~C., et al.\
1014: 2007, Publications of the Astronomical Society of Australia, 24, 1
1015:
1016: \bibitem[Kinney et al.(1996)]{1996ApJ...467...38K} Kinney, A.~L., Calzetti,
1017: D., Bohlin, R.~C., McQuade, K., Storchi-Bergmann, T., \& Schmitt, H.~R.\
1018: 1996, ApJ, 467, 38
1019:
1020: \bibitem[Kochanek(1992)]{1992ApJ...384....1K} Kochanek, C.~S.\ 1992, ApJ,
1021: 384, 1
1022:
1023: \bibitem[Kochanek et al.(1997)]{1997ApJ...479..678K} Kochanek, C.~S.,
1024: Falco, E.~E., Schild, R., Dobrzycki, A., Engels, D., \& Hagen, H.-J.\ 1997,
1025: ApJ, 479, 678
1026:
1027: \bibitem[Kochanek et al.(1999)]{1999ApJ...510..590K} Kochanek, C.~S.,
1028: Falco, E.~E., \& Mu{\~n}oz, J.~A.\ 1999, ApJ, 510, 590
1029:
1030: \bibitem[Kochanek et al.(2000)]{2000ApJ...543..131K} Kochanek, C.~S., et
1031: al.\ 2000, ApJ, 543, 131
1032:
1033: \bibitem[Kochanek et al.(2006)]{2006ApJ...637L..73K} Kochanek, C.~S.,
1034: Mochejska, B., Morgan, N.~D., \& Stanek, K.~Z.\ 2006, ApJL, 637, L73
1035:
1036: \bibitem[Kochanek et al.(2007)]{2006astro.ph..9112K} Kochanek, C.~S., Dai,
1037: X., Morgan, C., Morgan, N., Poindexter, S., \& Chartas, G.\ 2006, ArXiv
1038: Astrophysics e-prints, arXiv:astro-ph/0609112
1039:
1040: \bibitem[Koopmans \& Treu(2003)]{2003ApJ...583..606K} Koopmans, L.~V.~E.,
1041: \& Treu, T.\ 2003, ApJ, 583, 606
1042:
1043: \bibitem[Lacy et al.(2002)]{2002AJ....123.2925L} Lacy, M., Gregg, M.,
1044: Becker, R.~H., White, R.~L., Glikman, E., Helfand, D., \& Winn, J.~N.\
1045: 2002, AJ, 123, 2925
1046:
1047: \bibitem[Landolt(1992)]{1992AJ....104..340L} Landolt, A.~U.\ 1992, AJ,
1048: 104, 340
1049:
1050: \bibitem[Lawrence et al.(2006)]{2006astro.ph..4426L} Lawrence, A., et al.\
1051: 2006, astro-ph/0604426
1052:
1053: \bibitem[Linder(2004)]{2004PhRvD..70d3534L} Linder, E.~V.\ 2004, PhRvD, 70,
1054: 043534
1055:
1056: \bibitem[Mannucci et al.(2001)]{2001MNRAS.326..745M} Mannucci, F., Basile,
1057: F., Poggianti, B.~M., Cimatti, A., Daddi, E., Pozzetti, L., \& Vanzi, L.\
1058: 2001, MNRAS, 326, 745
1059:
1060: \bibitem[Maoz et al.(1993)]{1993ApJ...402...69M} Maoz, D., Bahcall, J.~N.,
1061: Doxsey, R., Schneider, D.~P., Bahcall, N.~A., Lahav, O., \& Yanny, B.\
1062: 1993, ApJ, 402, 69
1063:
1064: \bibitem[Menou et al.(2001)]{2001ApJ...561..645M} Menou, K., et al.\ 2001,
1065: ApJ, 561, 645
1066:
1067: \bibitem[Miyazaki et al.(2002)]{2002PASJ...54..833M} Miyazaki, S., et al.\
1068: 2002, PASJ, 54, 833
1069:
1070: \bibitem[Moro \& Munari(2000)]{2000A&AS..147..361M} Moro, D., \& Munari,
1071: U.\ 2000, A\&AS, 147, 361
1072:
1073: \bibitem[Mortonson et al.(2005)]{2005ApJ...628..594M} Mortonson, M.~J.,
1074: Schechter, P.~L., \& Wambsganss, J.\ 2005, ApJ, 628, 594
1075:
1076: \bibitem[Myers et al.(2003)]{2003MNRAS.341....1M} Myers, S.~T., et al.\
1077: 2003, MNRAS, 341, 1
1078:
1079: \bibitem[Ofek et al.(2002)]{2002MNRAS.337.1163O} Ofek, E.~O., Rix, H.-W.,
1080: Maoz, D., \& Prada, F.\ 2002, MNRAS, 337, 1163
1081:
1082: \bibitem[Ofek \& Maoz(2003)]{2003ApJ...594..101O} Ofek, E.~O., \& Maoz, D.\
1083: 2003, ApJ, 594, 101
1084:
1085: \bibitem[Ofek et al.(2003)]{2003MNRAS.343..639O} Ofek, E.~O., Rix, H.-W.,
1086: \& Maoz, D.\ 2003, MNRAS, 343, 639
1087:
1088: \bibitem[Ofek et al.(2006)]{2006ApJ...641...70O} Ofek, E.~O., Maoz, D.,
1089: Rix, H.-W., Kochanek, C.~S., \& Falco, E.~E.\ 2006, ApJ, 641, 70
1090:
1091: \bibitem[Oguri(2007)]{2007ApJ...660....1O} Oguri, M.\ 2007, ApJ, 660, 1
1092:
1093: \bibitem[Oguri et al.(2004)]{2004PASJ...56..399O} Oguri, M., et al.\ 2004,
1094: PASJ, 56, 399
1095:
1096: \bibitem[Oguri et al.(2005)]{2005ApJ...622..106O} Oguri, M., et al.\ 2005,
1097: ApJ, 622, 106
1098:
1099: \bibitem[Oguri et al.(2006)]{2006AJ....132..999O} Oguri, M., et al.\ 2006,
1100: AJ, 132, 999
1101:
1102: \bibitem[Oke et al.(1995)]{1995PASP..107..375O} Oke, J.~B., et al.\ 1995,
1103: PASP, 107, 375
1104:
1105: \bibitem[Oscoz et al.(1997)]{1997ApJ...491L...7O} Oscoz, A., Serra-Ricart,
1106: M., Mediavilla, E., Buitrago, J., \& Goicoechea, L.~J.\ 1997, ApJL, 491,
1107: L7
1108:
1109: \bibitem[Peng et al.(2002)]{2002AJ....124..266P} Peng, C.~Y., Ho, L.~C.,
1110: Impey, C.~D., \& Rix, H.-W.\ 2002, AJ, 124, 266
1111:
1112: \bibitem[Pindor et al.(2004)]{2004AJ....127.1318P} Pindor, B., et al.\
1113: 2004, AJ, 127, 1318
1114:
1115: \bibitem[Poindexter et al.(2007)]{2007ApJ...660..146P} Poindexter, S.,
1116: Morgan, N., Kochanek, C.~S., \& Falco, E.~E.\ 2007, ApJ, 660, 146
1117:
1118: \bibitem[Reimers et al.(2002)]{2002A&A...382L..26R} Reimers, D., Hagen,
1119: H.-J., Baade, R., Lopez, S., \& Tytler, D.\ 2002, A\&A, 382, L26
1120:
1121: \bibitem[Richards et al.(2004)]{2004AJ....127.1305R} Richards, G.~T., et
1122: al.\ 2004, AJ, 127, 1305
1123:
1124: %\bibitem[Richards et al.(2006)]{2006AJ....131...49R} Richards, G.~T., et
1125: %al.\ 2006, AJ, 131, 49
1126:
1127: \bibitem[Rusin et al.(2002)]{2002MNRAS.330..205R} Rusin, D., Norbury, M.,
1128: Biggs, A.~D., Marlow, D.~R., Jackson, N.~J., Browne, I.~W.~A., Wilkinson,
1129: P.~N., \& Myers, S.~T.\ 2002, MNRAS, 330, 205
1130:
1131: \bibitem[Rusin \& Kochanek(2005)]{2005ApJ...623..666R} Rusin, D., \&
1132: Kochanek, C.~S.\ 2005, ApJ, 623, 666
1133:
1134: \bibitem[Schechter \& Wambsganss(2002)]{2002ApJ...580..685S} Schechter,
1135: P.~L., \& Wambsganss, J.\ 2002, ApJ, 580, 685
1136:
1137: \bibitem[Schlegel et al.(1998)]{1998ApJ...500..525S} Schlegel, D.~J.,
1138: Finkbeiner, D.~P., \& Davis, M.\ 1998, ApJ, 500, 525
1139:
1140: \bibitem[Schneider et al.(2007)]{2007AJ....134..102S} Schneider, D.~P., et
1141: al.\ 2007, AJ, 134, 102
1142:
1143: \bibitem[Spergel et al.(2007)]{2007ApJS..170..377S} Spergel, D.~N., et al.\
1144: 2007, ApJS, 170, 377
1145:
1146: \bibitem[Treu \& Koopmans(2004)]{2004ApJ...611..739T} Treu, T., \&
1147: Koopmans, L.~V.~E.\ 2004, ApJ, 611, 739
1148:
1149: \bibitem[Treu et al.(2006)]{2006ApJ...640..662T} Treu, T., Koopmans, L.~V.,
1150: Bolton, A.~S., Burles, S., \& Moustakas, L.~A.\ 2006, ApJ, 640, 662
1151:
1152: \bibitem[Turatto et al.(2003)]{2003fthp.conf..200T} Turatto, M., Benetti,
1153: S., \& Cappellaro, E.\ 2003, From Twilight to Highlight: The Physics of
1154: Supernovae, 200
1155:
1156: \bibitem[Tyson(2005)]{2005ASPC..339...95T} Tyson, A.\ 2005, Observing Dark
1157: Energy, 339, 95
1158:
1159: \bibitem[Vanden Berk et al.(2001)]{2001AJ....122..549V} Vanden Berk, D.~E.,
1160: et al.\ 2001, AJ, 122, 549
1161:
1162: \bibitem[Wambsganss et al.(2000)]{2000A&A...362L..37W} Wambsganss, J.,
1163: Schmidt, R.~W., Colley, W., Kundi{\'c}, T., \& Turner, E.~L.\ 2000, A\&A,
1164: 362, L37
1165:
1166: \bibitem[Wisotzki et al.(1993)]{1993A&A...278L..15W} Wisotzki, L., Koehler,
1167: T., Kayser, R., \& Reimers, D.\ 1993, A\&A, 278, L15
1168:
1169: %\bibitem[Witt et al.(2000)]{2000ApJ...544...98W} Witt, H.~J., Mao, S., \&
1170: %Keeton, C.~R.\ 2000, ApJ, 544, 98
1171:
1172: \bibitem[Wyithe et al.(2003)]{2003ApJ...583...58W} Wyithe, J.~S.~B., Winn,
1173: J.~N., \& Rusin, D.\ 2003, ApJ, 583, 58
1174:
1175: \bibitem[York et al.(2000)]{2000AJ....120.1579Y} York, D.~G., et al.\ 2000,
1176: AJ, 120, 1579
1177:
1178: \end{thebibliography}
1179:
1180:
1181: \end{document}
1182:
1183:
1184: