1: % Revised MS 7/14/06
2: \documentclass[10pt,preprint]{aastex}%
3: \begin{document}
4:
5: \title{The Spectral Energy Distribution of the High-Z Blazar Q0906+6930}
6: \author{Roger W. Romani}
7: \affil{Department of Physics, Stanford University, Stanford, CA 94305}
8: \email{rwr@astro.stanford.edu}
9:
10: \begin{abstract}
11: We describe further observations of QSO J0906+6930, a z=5.48 blazar
12: likely to be detected in $\gamma$-rays. New radio and X-ray data place
13: significant constraints on any kpc-scale extension of the VLBA-detected jet.
14: Improved optical spectroscopy detects absorption from an intervening galaxy at
15: z=1.849 and raise the possibility that this distant, bright source is
16: lensed. We combine the new data into an improved SED for the blazar
17: core and comment on the Compton keV-GeV flux component.
18: \end{abstract}
19:
20: \keywords{galaxies: jets -- quasars: general}
21:
22: \section{Introduction}
23:
24: Q0906+6930 (=GB6 J0906+6930=SRM J090630.74+693030.8)
25: was discovered in a survey of radio-bright, flat spectrum
26: sources chosen to be like the EGRET blazars \citep{srm03,set05}. Follow-up
27: observations confirmed a large $z\approx 5.5$ redshift and found evidence
28: for a very compact pc-scale jet with the VLBA \citep{ret04}. The source is
29: quite radio-loud with an $R = f_{\rm 5\,GHz}/f_{\rm 440nm}({\rm rest})
30: \approx 500-1000$, supporting the blazar interpretation. These data
31: showed several peculiar properties. For example, the radio spectrum appeared
32: to steepen above 10\,GHz, while the jet component still appeared to be inverted.
33: Also, the source shows a relatively large $1350$\AA\, continuum flux and
34: large kinematic widths for the emission lines. These suggest a large
35: $\sim 2 \times 10^9M_\odot$ black hole mass [note that an error in the
36: $\lambda F_\lambda$ luminosity quoted in \citet{ret04} implied
37: $M\ge 10^{10}M_\odot$].
38: While this source was, at best, a weak background enhancement in the
39: {\it EGRET} survey, the prospects for detection with GLAST seem quite strong.
40: This is particularly interesting as observations of a cut-off in the blazar
41: spectrum above $\sim 10$\,GeV can be used to probe absorption by
42: light produced at the peak of star formation \citep{crr04}.
43:
44: We report here on further observations of Q0906+6930, which support
45: its identification as a high-z blazar, test the nature of the jet component
46: and probe it's status as a high mass, high luminosity source.
47:
48: \section{New Observations}
49:
50: Several radio-loud QSO at high redshift show kpc-scale jets
51: so our first objective was to constrain such arcsecond-scale emission.
52: At the blazar redshift, $1^{\prime\prime}= 6.19$\,kpc (for an
53: % 8.529 at z=1.849
54: $\Omega_m= 0.24$, $H_0= 73$\,km/s/Mpc flat cosmology, assumed throughout;
55: Spergel et al. 2006). Perhaps the most convincing case of such jet emission
56: is still that of QSO GB 1508+5714 at z=4.3, which is
57: detected both in the X-ray \citep{sie03} and radio \citep{che04} bands.
58: Its large observed $f_X/f_R \sim 10^2$ supports a scenario where the
59: jet electrons Compton up-scatter CMB seed photons, which have a large
60: increase in energy density, $u \propto (1+z)^4$ at high redshift \citep{sch02}.
61:
62: \subsection{X-ray imaging}
63:
64: We obtained a 30\,ks {\it Chandra} observation of the blazar
65: on 2005 July 1 positioned at the best focus of the ACIS-S3
66: backside-illuminated chip. The observations were standard full-frame
67: 3.2s TE exposures in very faint (VF) mode.
68: No strong background flaring was seen during this observation so that
69: all of the data could be included in the analysis.
70: The data were processed using standard algorithms in CIAO version
71: 3.2 and CALDB version 3.0.0.
72: The blazar was clearly detected, providing $\sim 435$ counts, at
73: a position coincident with the radio core. With this modest flux,
74: pile-up is under 1\%, and so we can ignore it in the subsequent analysis.
75: The X-ray image is consistent with a PSF computed for the blazar
76: spectrum; the largest excess (at $\sim 1.5^{\prime\prime}$, position
77: angle $\sim 115^\circ$) contains 4 photons and is not significant.
78: Thus, on $\sim 1-10$ arcsecond scales, any
79: extended (jet) emission must have $\le 1$\% of the core flux.
80: A number of coronal emission field stars and some unidentified
81: objects (presumably background AGN) were detected in this integration, but
82: none showed notable peculiarities, so we do not discuss these further.
83:
84: \begin{figure}[!h]
85: %\includegraphics[scale=0.45]{Q0906SEDfig1.eps}
86: %\plotone{Q0906SED_fig1.eps}
87: %\plotfiddle{Q0906SEDfig1.pdf}{.1truecm}{0}{250}{250}{100}{0}
88: %\plotfiddle{Fig1b.pdf}{.1truecm}{0}{100}{100}{300}{-300}
89: \plottwo{fig1a.eps}{fig1b.eps}
90: %
91: %\includegraphics[clip=true,scale=0.25,trim=-650 -550 0 0]{Fig1b.pdf}
92: %\includegraphics[scale=0.6,trim=450 0 0 0]{Q0906SEDfig1.pdf}
93: \figcaption{\small
94: {\it CXO} ACIS-S3 spectrum of Q0906+6930, with the best-fit absorbed
95: power-law model. The second panel shows the 0.5-8\,keV {\it CXO} image.
96: }
97: \end{figure}
98:
99: The X-ray spectrum (figure 1) is adequately described by a simple
100: absorbed power-law,
101: with a low absorption $N_H = 5.4\pm2 \times 10^{20} {\rm cm^{-2}}$,
102: a photon index $\Gamma = 1.6 \pm 0.1$ and an unabsorbed flux
103: $f_X (0.2-10\,{\rm keV})) = 1.47\pm 0.25 \times 10^{-13}
104: {\rm erg\,cm^{-2}\,s^{-1}}$.
105: Although the data show a small excess to the best fit
106: power-law model around 1\,keV, an attempt to fit a red-shifted Fe line
107: produced no significant detection. The upper limit
108: to the line flux is $\sim 1.5 \times 10^{-15}{\rm erg\,cm^{-2}\,s^{-1}}$,
109: or $\le 1$\% of the continuum flux. These spectral parameters are quite
110: typical of lower redshift blazars. The measured absorption is consistent
111: with the Galactic column in this direction, $\sim 4.4 \times 10^{20} {\rm cm^{-2}}$ from
112: the COLDEN tool. We can also compare the X-ray properties with
113: those of a set of radio-loud high-redshift QSO recently reported by
114: \citet{lopet06}. Our observed $\Gamma$ and inferred
115: $\alpha_{OX} [250\,{\rm nm/2\,keV}] (\rm rest) \approx -1.3$, are in excellent
116: agreement with the objects measured in this paper. We conclude that this
117: is a fairly typical radio-bright quasar at high redshift.
118:
119: \subsection{Radio imaging}
120:
121: To search for a larger scale counterpart of the VLBI jet reported
122: in \citet{ret04}, we also obtained VLA images with the A array at
123: 1.4-, 4.8- and 8.4\,GHz on 2004, October 7. The on-source times were
124: 10m, 30min and 50min
125: respectively. These data were calibrated and mapped using
126: standard AIPS routines and were found to be adequately described by
127: a simple point source model. The final map noise was within a factor
128: of $\sim 2$ of the thermal noise in each case. Measuring the rms background
129: in an annulus between the beam size and $\sim 5^{\prime\prime}$,
130: we obtain limits on any jet component (Table 1). The largest
131: map fluxes in the searched regions are consistent with statistical
132: fluctuations for these rms vales, so we can take upper limits on
133: the jet flux along the position angle of the mas jet to be $\sim 2\times$
134: the local rms, while the limits on any extended component are about
135: two times larger. For the X-band image this search extends within
136: $\sim 0.4^{\prime\prime}$ of the blazar core, while for Q-band in
137: the B-array, we searched down to $0.1^{\prime\prime}$.
138:
139: Since no extended emission is seen in the A-array data,
140: we used follow-on B-array observations on 2005 March 04 to obtain a
141: six frequency, single epoch radio spectrum. The on-source total scan times
142: are listed in Table 1. The data were calibrated by observations of
143: 0542+498 (=3C 147) at each frequency. For all measurements, the statistical
144: error in the flux density estimate was small. We add this in quadrature
145: to the flux scale uncertainties described in the VLA calibrator manual
146: (2\% at L, C and X and 5\% at
147: U, K and Q) to obtain the flux density errors in table 1. For
148: K and Q residual pointing errors and uncorrected atmospheric absorption
149: may contribute additional error.
150: In all cases a single unresolved component fit provided
151: a satisfactory description of the Q0906+6930 data. In Table 1, we also
152: give RMS fluctuation levels in an annulus about the point source. The local rms
153: limits to extended flux for the three lower frequency bands are from the
154: deeper, higher-resolution A-array images.
155: %\clearpage
156:
157: \begin{table}[!ht]
158: \caption{Radio Core Fluxes and Jet Bounds}
159: \medskip
160: \tabletypesize=\small
161: \begin{tabular}{llllll}
162: Band & Freq(GHz) &On-Source time (min)& Core Flux (mJy) & Error (mJy) & Jet Annulus RMS (mJy)\\
163: \hline
164: L & 1.44 & 8 & 92 & 2.0 & 0.10$^\dagger$\\ %1mJy, 2% calibration
165: C & 4.88 & 8 & 114 & 2.4 & 0.05$^\dagger$\\ % 0.7mJy 2% cal
166: X & 8.44 & 12 & 136 & 3.1 & 0.13$^\dagger$\\ %1.5mJy 2% cal
167: U & 14.96 & 24 & 129 & 6.6 & 0.23\\% 1.2mJy + 5% Re-fit got 129+/-0.2
168: K & 22.49 & 15 & 83 & 4.2 & 0.22\\%0.4mJy + 5% 94.6+/-0.35, 0.22
169: Q & 43.33 & 30 & 43 & 2.2 & 0.26\\% 0.3mJy + 5% 44.4, 0.26
170: \end{tabular}
171: \leftline{$^\dagger$ RMS from 2004/10/04 A-array maps}
172: \label{radio_flux}
173: \end{table}
174:
175: \subsection{Optical Spectroscopy and Imaging}
176:
177: To further investigate the absorbed optical spectrum reported
178: in \citet{ret04}, we obtained higher resolution spectra using the
179: Hobby$\bullet$Eberly telescope \citep[HET;][]{ram98}
180: Marcario Low Resolution Spectrograph \citep[LRS;][]{hill98} using
181: the G3 VPH grating, which provides 1.88\AA/pixel dispersion over the range
182: $\lambda\lambda$6250\AA\, - 9100\AA. The integrations used a $1.5^{\prime\prime}$\
183: slit, giving an effective resolution of
184: $5.5$\AA. Standard queue observations were made on 2005 February 9 and 14
185: and 2005 May 9-10; conditions were generally sub-par with poor transparency
186: or variable seeing. After standard IRAF processing and combination using
187: the relative S/N, seven exposures totaling 8600s
188: of integration were assembled to produce the spectrum in figure 2.
189: Here the new G3 spectrum is plotted with a heavy line over the lower resolution
190: G1 spectrum of \citet{ret04}. The excellent correspondence between the
191: fine structure in these spectra, especially in the Ly$\alpha$ forest region,
192: attests to the stable calibration. The combined spectrum has a S/N per
193: resolution element of $\sim 16$ in the unabsorbed continuum.
194:
195: \begin{figure}
196: %\plotone{Q0906SED_fig2.eps}
197: \plotfiddle{fig2.eps}{.0truecm}{90}{650}{420}{0}{00}
198: \figcaption{\small
199: HET/LRS G3 spectrum of Q0906+6930, over plotted on lower resolution
200: data with a wider wavelength coverage. The inset shows the G3 data alone,
201: with two metal line absorption systems indicated.
202: }
203: \end{figure}
204:
205: These higher resolution data allow us to check the reality of
206: absorption-line systems suggested in the G1 spectrum. Two metal line
207: systems can be seen in the inset of figure 2, which shows the spectrum redward
208: of Ly$\alpha$. In particular, we confirm a likely C IV doublet at $z=4.698$,
209: with a rest equivalent width (REW) for each component of $W_0=0.6\pm 0.13$\AA.
210: This system has an associated, apparently damped, Ly$\alpha$ absorption
211: and a minimum at Ly$\beta$. More interesting is an apparent strong Mg II
212: absorption at $z=1.849$ which was confused by possible N\,V emission in the
213: lower resolution data. The REW of the 2795\AA\, line is $1.9\pm0.07$\AA\,
214: and the line spacing and doublet ratio $R=1.2$ support this identification.
215: In our restricted spectral range no strong confirming line is expected.
216:
217: The presence of intervening absorption line systems raises the
218: possibility that the blazar is magnified by gravitational
219: lensing, as discussed below. To constrain this hypothesis, we examined
220: the limited imaging data. During HET queue spectroscopy, 10-30s
221: $\sim$white light set-up images are generally recorded through a 515nm
222: long-pass filter. Six of these images had usable ($\le 1.7^{\prime\prime}$)
223: image quality; these were S/N weighted and combined to produce
224: a final image with $1.6^{\prime\prime}$ FWHM. This was used to
225: get an approximate limit on any extended source within $5^{\prime\prime}$
226: of Q0907+6930 of $R\sim23.1$. Imaging reconnaissance was also obtained
227: using the MDM 2.4m RETROCAM on 2006 February 4, courtesy of J. Marshall
228: and D. DePoy. Here $6\times 300$s $r$ images ($\sim 1.3^{\prime\prime}$ FHWM)
229: were combined. Although Ly$\alpha$ forest absorption should suppress
230: the quasar relative to the foreground galaxy in the $r$ band, in the
231: combined image the blazar was again stellar, with an apparent magnitude
232: of $r=21.9$. We derive an upper limit of $r>23.3$ on any
233: nearby galactic companion.
234:
235:
236: \section{Spectral Energy Distribution}
237:
238: We have assembled these new flux measurements into a (non-instantaneous)
239: spectral energy distribution (SED), figure 3. The plot shows the 04/03/05
240: simultaneous radio fluxes. We do not see any strong evidence of radio
241: variability, but it should be noted that the X-ray flux (shown with a dashed
242: lines for the continuum uncertainty range) was observed $\sim 100$d later.
243: The optical spectroscopy measurements span several years, although again we
244: do not see variability. With a rather small range of unabsorbed continuum
245: available, the optical spectral index is quite uncertain (dashed lines, extended a
246: a decade for visibility). This situation would be greatly helped by near-IR
247: spectroscopy. Finally, the {\it EGRET} upper limit is, of course, from
248: 10-15 years prior to the current observations, and blazars are known to be
249: highly variable in the $\gamma$-ray band. The low significance detection during
250: one viewing period \citep{ret04} should be seen as consistent with the
251: mission-averaged upper limit.
252:
253: As in \citet{ret04} we compare the new SED with that of the
254: well-studied EGRET blazar 3C 279 [inset: SED from period P5b \citep{har01},
255: red-shifted to z=5.48], and with synchro-Compton models. The Q0906+6930
256: spectrum is sparse, but notably shows a significantly ($\ge 3\times$) more
257: luminous optical spectrum. In contrast to 3C 279 the optical emission shows
258: high equivalent-width lines from the broad-line region (BLR),
259: and is thus not dominated by synchrotron continuum. For the Compton components
260: of the SED, our new X-ray detection confirms a rising $\nu F_\nu$
261: spectrum, likely up-scattered by the synchrotron-emitting electrons.
262: This is somewhat ($\sim 50$\%) fainter than the 3C 279 X-ray luminosity,
263: which is reasonable in light of the lower radio synchrotron luminosity.
264: However, the possible GeV detection would require an external Compton component
265: $\sim 30\times$ brighter than that of 3C 279.
266:
267: By computing representative synchro-Compton models, we see that the brighter
268: disk emission and a luminous $\gamma$-ray Compton component
269: may be related. In the inset we show a model for 3C 279 computed
270: from a code kindly supplied by M. B\"ottcher. This code computes synchrotron
271: and Compton emission in an evolving cylindrical jet outflow. It also models a host
272: accretion disk and computes the Comptonized flux from these soft disk photons,
273: both directly from the disk and disk photons scattered from the broad line
274: region. The code is described in \citet{boet97} and \citet{bb00}. To generate the 3C 279
275: model, we use explicitly the parameters listed in \citet{har01} for this
276: epoch. The `jet' is a $6\times 10^{16}$\,cm radius blob injected 0.025\,pc
277: from the $1.5\times 10^8M_\odot$ hole, moving cylindrically at $2^\circ$
278: to the line of sight with
279: bulk $\Gamma=13$. The jet contains 35${\rm cm^{-3}}$ each of $e^\pm$,
280: with a power-law energy distribution of index $p=3.0$, extending from
281: $\gamma_1=6\times 10^2$ to $\gamma_2=1\times 10^5$ and 1.5\,G of
282: magnetic field. The hole is surrounded
283: by an accretion disk of luminosity $10^{46}{\rm erg/s}$ and a broad-line
284: region extending from 0.1-0.4\,pc. To reproduce the model shown in
285: \citep{har01}, it was necessary to increase the BLR Thomson depth to $\tau=0.06$.
286: These parameters gave the total SED curve shown in the inset.
287:
288: We know that Q0906+6930 has a higher mass hole ($\sim 2 \times 10^9M_\odot$)
289: and a brighter accretion disk ($L_D \ge 3 \times 10^{46}{\rm erg/s}$). We
290: made these two changes to the parameters and re-computed the SED, as shown
291: in the main part of Figure 3. Under the full SED we show the individual
292: components: (from low to high frequency) the jet synchrotron emission,
293: the disk thermal emission, the jet self-Compton emission, the external Compton
294: emission from the disk and the Comptonized emission from disk photons
295: scattered off the broad-line
296: region. The spectra were computed at $z=0.538$ and then the 3C 279 data and
297: the two models were red-shifted to z=5.48 using concordance flat cosmology
298: parameters. These models illustrate that with a higher mass hole and
299: a brighter, cooler disk, brighter external Compton ($\gamma$-ray) emission
300: can be expected. Of course, such models have many parameters and fits to more
301: densely sampled SEDs in a variety of flare states are needed for any predictive
302: power.
303:
304: A standard definition of radio-loudness is the ratio
305: $R=f_{\rm 5\,GHz}/f_{\rm 440nm}({\rm rest})$ \citep{ket89}. Although we do not
306: have direct
307: observations at $770$\,MHz or $2.9\mu$, we can extrapolate the observed
308: optical and radio spectra to estimate $R\approx 540-970$ where the range
309: %f_0.77GHz = 8.5e-25, f_2.8mu = 8.8e-28 to 1.6 e-27
310: is due to the rather uncertain optical continuum spectrum.
311:
312: We should note that both the 3C 279 and the Q0906+6930 models
313: safely under-predict the radio spectrum. This is because the model
314: simulates only the base of the jet emission, omitting larger
315: scales where cooling, expansion and injection of additional $e^\pm$ populations
316: produce the cm-wavelength radio components. In particular, the
317: local maximum in the Q0906+6930 radio spectrum at $\sim 15$GHz ($\sim 100$GHz
318: rest-frame) suggests a distinct radio jet component. It should be noted that
319: the 43\,GHz flux measured here (on kpc scales) is in good agreement with the
320: {\it VLBA} core measurement. Thus at most 10\% percent of the radio flux lies
321: in the $\sim 5-10$\,pc resolved VLBA jet and the cm-wavelength radio component
322: is produced at sub-pc scales in the unresolved VLBA core.
323: It would be useful to check if the SED rises
324: toward a synchrotron peak in the far-IR/sub-mm region as expected.
325: IR photometry of the blazar will also be helpful in constraining the
326: level of the synchrotron SED and confirming the dominance of disk emission
327: in the optical.
328:
329: \begin{figure}[h!]
330: \plotone{fig3.eps}
331: %\plotfiddle{/home/rwr/Blazar/jets/Q0906SED.eps}{4.0truecm}{0}{50}{50}{0}{0}
332: \figcaption{\small Multiwavelength spectral energy distribution for Q0906+6930,
333: from non-simultaneous data (open circles).
334: The optical continuum flux is plotted with a line (short dash) extended to
335: show the slope. The X-ray spectrum (solid line) shows our {\it CXO}
336: continuum fit. For these
337: two components dashed lines show the range of spectral index uncertainty.
338: For the $\gamma$-ray band we show the brightest epoch {\it EGRET} flux,
339: the {\it EGRET} mission-averaged upper limit and the {\it GLAST} sensitivity
340: of the first year sky survey. A synchro-Compton model is plotted (see text).
341: The individual component curves (synchrotron emission, synchrotron self-Compton
342: emission, disk emission [dashed line] and Comptonized emission directly
343: from the disk and after scattering in the broad line region) are shown.
344: For comparison, the inset shows the SED of 3C 279 during period P5b \citet{har01},
345: shifted to z=5.48, along with its synchro-Compton model (see text).
346: }
347: \end{figure}
348:
349: \section{Conclusions}
350:
351: The new data generally confirm the SED picture described in
352: \citet{ret04} and show that at least the X-ray Compton component is present in
353: this blazar. The lack of any resolved kpc-scale jet component suggests
354: that all emission observed is from the core (or a compact pc-scale jet).
355: Also, this does not support the \citet{sch02} picture where large-scale
356: X-ray jets are produced by up-scatter of the CMB. In this model, the
357: $\propto (1+z)^4$ increase of the CMB energy density counters the
358: decreasing jet surface brightness; at $z>4$ jet emission is thus expected
359: to exceed
360: core emission. With $\le 1$\% of the core X-ray flux in a resolvable jets,
361: our results extend those of \citet{lopet06}, suggesting core photons,
362: rather than CMB photons dominate the flux up-scattered as Compton
363: X-ray emission.
364:
365: One peculiarity noted in \citet{ret04} was the relatively large optical and radio
366: flux; we noted the possibility of lensing of the QSO core. This should be
367: re-visited, since the strong intervening metal line absorbers suggest
368: galactic-scale
369: masses along the line-of-sight. The most important is the MgII system at
370: z=1.849, which at $W_0=1.9$\AA, likely represents a classical $\sim L_\ast$
371: spiral galaxy \citep{chet05} which should be a Lyman Limit System (LSS)
372: absorber, within an impact parameter
373: $$
374: R = 38 h^{-1}(L_K/L_K^\ast)^{0.15} {\rm kpc},
375: $$
376: for an $L_\ast$ galaxy at the absorber redshift, i.e. within $6.2^{\prime\prime}$
377: for our assumed cosmology. Our limited direct imaging does not detect such a
378: galaxy, but the results are not very constraining. At this $z$, the cosmic age is
379: $\sim 3.6$\,Gyr and an $L_\ast$ Sa formed at z=5 in our cosmology would have
380: $r\sim 23.9$ for a \citet{bc93} model,
381: %http://qso.as.arizona.edu/~ckeeton/mag_evol/index.html
382: i.e. about $2\times$ fainter than our imaging $r\sim 23.3$ limit. Sbc types
383: would be even fainter.
384:
385: Can such a galaxy lens our blazar? The expected image angular scale
386: for a lens mass $M$ is
387: $$
388: \theta_0 = \left [ 4 D_{LS}G\,M/(D_S D_L c^2)\right ]^{1/2}
389: $$
390: where one uses angular diameter distances and $D_{LS} = D_s-(1+z_L)D_L/(1+z_s)$.
391: The geometry is favorable for lensing and the lens scale is then
392: $1.35 \times 10^{-6}M^{1/2}$arcsec with $M$ in $M_\odot$. We see no evidence for a
393: double image of the blazar on arcsecond scales, either in the optical or
394: the radio, so this precludes classical macro lensing by an $L_\ast$ galaxy.
395: Indeed, the core is unresolved in our VLBI maps, where the half-power beam width
396: at 43\,GHz was 0.55$\times 0.30$\,mas, so we can place an upper limit on
397: an effective lens mass of $M\le 1.4 \times 10^5 M_\odot$. The lack of obvious
398: variability on year time-scales suggests a not very constraining lower
399: mass limit of
400: $M\ge 10^{-3} v_{300}^2 M_\odot$ for a lens galaxy velocity dispersion
401: of $\sim 300\,v_{300}$\,km/s. We may infer a slightly tighter limit if
402: we note that the large BLR equivalent widths imply that the broad line
403: region must be lensed along with the core. Naively applying the continuum
404: BLR radius correlation of \citet{kasp05} to our optical continuum luminosity
405: $\lambda F_\lambda |_{5100\AA} \approx 3 \times 10^{46} {\rm erg/s}$
406: we conclude
407: $$
408: R_{BLR} \approx 5.8 \times 10^{16}
409: \left ( \lambda F_\lambda |_{5100\AA}/10^{44}{\rm erg/s} \right )^{0.69}
410: \approx 2.9 \times 10^{18}{\rm cm}.
411: $$
412: This in turn implies a minimum lens mass $M\ge 10^{-2}M_\odot$. We infer
413: that macro-lensing does not amplify Q0906+6930, but that micro- or milli-lensing
414: are possible. On the other hand, it should be noted that the probability
415: of an unrelated intervening galaxy is not small. Since $dN/dz(W_0>1.9\AA) \approx
416: 0.13$ over the redshift range covered by our G3 spectrum \citep{net05}, we
417: expect 0.06 such absorbers. Detection of one MgII systems is thus not very improbable
418: and so we cannot infer from this that lensing amplification has brightened
419: Q0906+6930.
420:
421: We conclude by re-iterating that our new SED measurements
422: support the picture of Q0906+6930 as a blazar with a bright $\gamma$-ray
423: Compton component emitting at z=5.48. This Compton radiation must
424: traverse the peak of star formation at $z\approx 2-3$, where
425: optical/UV emission can attenuate the $>$10\,GeV photons, just as the IR
426: background can attenuate the TeV emission of blazars at lower $z$,
427: e.g. \citet{dka05} and references therein.
428: Of course, there is also
429: attenuation from the host photon field and, in principle, from the intervening
430: z=1.89 galaxy. These contributions will make it difficult to extract
431: constraints on the extragalactic background photon field from this (or any one)
432: object. However, statistical studies of high red-shift blazars with GLAST
433: \citep{crr04}, should still be able to extract global constraints on the
434: extragalactic background light and its evolution.
435:
436:
437: \acknowledgments
438:
439: {\small
440: We thank the referee for a careful reading. We also thank many
441: colleagues for assistance with the collection of the SED data: in the radio
442: Lincoln Greenhill and Greg Taylor, in the optical Steve Healey and in the
443: X-ray Stephen Ng. MDM images were kindly obtained by Darren DePoy and
444: Jennifer Marshall.
445: This work was supported in part by NASA grant G05-6101X issued by the Chandra
446: X-ray Observatory Center, which is operated by the Smithsonian Astrophysical
447: Observatory for and on behalf of the National Aeronautics Space Administration
448: under contract NAS8-03060.
449:
450: The Hobby-Eberly Telescope (HET) is a joint project of the University
451: of Texas at Austin, the Pennsylvania State University, Stanford
452: University, Ludwig-Maximilians-Universit\"at M\"unchen, and
453: Georg-August-Universit\"at G\"ottingen. The HET is named in honor of
454: its principal benefactors, William P. Hobby and Robert E. Eberly.
455: The Marcario Low Resolution Spectrograph is named for Mike Marcario
456: of High Lonesome Optics who fabricated several optics for the instrument
457: but died before its completion. The LRS is a joint project of the
458: Hobby-Eberly Telescope partnership and the Instituto de Astronomia de
459: la Universidad Nacional Autonoma de Mexico.
460: The National Radio Astronomy Observatory is a facility of the
461: National Science Foundation operated under cooperative agreement
462: by Associated Universities, Inc.
463: }
464:
465: \begin{thebibliography}{}
466:
467: %\bibitem[Boettcher(2003)]{boet03}Böttcher, M., 2003. Proc. XXII Moriond Astrophysics Meeting, 151
468:
469: \bibitem[B\"ottcher, et al.(1997)]{boet97}B\"ottcher, M., Mause, H. \&
470: Schlickeiser, R. 1997, A\&A, 324, 39, A\&A, 324, 395
471:
472: \bibitem[B\"ottcher \& Bloom(2000)]{bb00}B\"ottcher, M., \& Bloom, S.D. 2000, AJ, 119, 469.
473:
474: \bibitem[Bruzual \& Charlot(1993)]{bc93}Bruzual, G. \& Charlot, S. 1993, ApJ,
475: 405, 538.
476:
477: \bibitem[Chen, Reyes \& Ritz(2004)]{crr04}Chen, A., Reyes, L. \& Ritz, S. 2004,
478: ApJ, 608, 686
479:
480: \bibitem[Cheung(2004)]{che04}Cheung, C.C. 2004, \apj, 600, L23
481:
482: \bibitem[Churchill et al.(2005)]{chet05}Churchill, C.W., Kacprzak, G.C. \& Steidel, C. C. 2005, Proc. IAU, 1, 24; astro-ph/0504392
483:
484: \bibitem[Dwek, Krennrich \& Arendt(2005)]{dka05}Dwek, E., Krennrich, F. \& Arendt, R.G. 2005, ApJ, 634, 155
485:
486: \bibitem[Hartman et al.(2001)]{har01}Hartman, R.C. et al. 2001, \apj, 553, 683
487:
488: \bibitem[Hill et al.(1998)]{hill98}Hill, G.J. et al. 1998, Proc. SPIE, 3355, 375
489:
490: %bibitem[Landt et al.(2001)]{lan01}Landt, H. et al. 2001, \mnras, 323, 757
491:
492: \bibitem[Kaspi et al.(2005)]{kasp05}Kaspi, S. et al. 2005, ApJ, 629, 1
493:
494: \bibitem[Kellerman et al.(1989)]{ket89}Kellerman, K.I. et al. 1989, AJ, 98, 1195
495:
496: %\bibitem[Krawczynski et al.(2004)]{kra04}Krawczynski, H., Hughes, S. B., Horan, D., et al., 2004, \apj, 601, 151
497:
498: \bibitem[Lopez et al.(2006)]{lopet06}Lopez, L.S. et al. 2006, AJ, 131, 1914
499:
500: %\bibitem[Marscher et al(2002)]{mar02}Marscher, A.P. et al 2002, ApJ 577, 85
501:
502: %bibitem[Madau \& Phinney(1996)]{mp96}Madau, P. \& Phinney, E.S. 1996, \apj, 456, 124
503:
504: %bibitem[Myers et al(2003)]{mye03}Myers, S. T. et al 2003, \mnras, 341, 1
505:
506: \bibitem[Nestor et al.(2005)]{net05}Nestor, D.B. et al 2005, ApJ 628, 637
507:
508: \bibitem[Ramsey et al.(1998)]{ram98}Ramsey, L.W. et al 1998, Proc. SPIE, 3355, 34
509:
510: \bibitem[Romani et al.(2004)]{ret04} Romani, R.W. et al 2004, ApJ 610, L9
511:
512: \bibitem[Schwartz(2002)]{sch02}Schwartz, D.A. 2002, \apj, L23
513:
514: \bibitem[Siemiginowska et al.(2003)]{sie03}Siemiginowska, A. et al 2003, \apj, 598, 15
515:
516: \bibitem[Sowards-Emmerd, et al.(2004)]{srmu04}Sowards-Emmerd, D.,
517: Romani, R. W., Michelson, P. F. \& Ulvestad, J. 2004, \apj, 609, 564.
518:
519: \bibitem[Sowards-Emmerd, et al.(2005)]{set05}Sowards-Emmerd, D.,
520: Romani, R. W., Michelson, P. F., Healey, S.E. \& Nolan, P. 2005, \apj, 626, 95.
521:
522: \bibitem[Sowards-Emmerd, Romani \& Michelson(2003)]{srm03}Sowards-Emmerd, D.,
523: Romani, R. W., \& Michelson, P. F. 2003, \apj, 590, 109
524:
525: \bibitem[Spergel, at al.(2006)]{set06}Spergel, D. et al 2006, ApJ, in press; astro-ph/0603449
526:
527: %bibitem[Urry \& Padovani(1995)]{up95}Urry, C.M. \& Padovani, P. 1995, \pasp, 107, 803
528:
529: \bibitem[Vestergaard(2002)]{ves02}Vestergaard, M. 2002, \apj, 571, 733
530:
531: \end{thebibliography}
532:
533: \end{document}
534:
535:
536: