1: \documentclass{XrU2005}
2: \usepackage{graphicx}
3:
4: \title{The High-Velocity Outflow of PG 1211+143: An Unbiased View Based on Several Observations}
5: \author[1,2]{S. Kaspi}
6: \author[1]{E. Behar}
7: \affil[1]{Department of Physics, Technion, Haifa 32000, Israel}
8: \affil[2]{School of Physics and Astronomy and the Wise
9: Observatory, Tel-Aviv University, Tel-Aviv 69978, Israel}
10:
11: \newcommand{\btx}{\textsc{Bib}\TeX}
12: \newcommand{\filename}{XrU2005}
13: \newcommand \kms {\ifmmode {\rm km\,s}^{-1} \else km\,s$^{-1}$\fi}
14: \newcommand \cmii {\hbox{cm$^{-2}$}}
15: \newcommand \ergs {\ifmmode {\rm erg\,s}^{-1} \else erg s$^{-1}$\fi}
16: \newcommand \ergcms {\ifmmode {\rm erg\,cm}^{-2}\,{\rm s}^{-1}
17: \else erg\,cm$^{-2}$\,s$^{-1}$\fi}
18: \newcommand \ergcmsA {\ifmmode{\rm erg\,cm}^{-2}\,{\rm s}^{-1}\,{\rm\AA}^{-1}
19: \else erg\,cm$^{-2}$\,s$^{-1}$\,\AA$^{-1}$\fi}
20: \newcommand \ergcmsHz {\ifmmode{\rm erg\,cm}^{-2}\,{\rm s}^{-1}\,{\rm Hz}^{-1}
21: \else erg\,cm$^{-2}$\,s$^{-1}$\,Hz$^{-1}$\fi}
22: \newcommand \phcms {\ifmmode {\rm ph\,cm}^{-2}\,{\rm s}^{-1}
23: \else ph\,cm$^{-2}$\,s$^{-1}$\fi}
24: \newcommand \phcmsA {\ifmmode {\rm ph\,cm}^{-2}\,{\rm s}^{-1}\,{\rm\AA}^{-1}
25: \else ph\,cm$^{-2}$\,s$^{-1}$\,\AA$^{-1}$\fi}
26:
27: \newcommand \pgt {PG\,1211+143}
28: \newcommand \xmm {{\it XMM-Newton}}
29: \newcommand \rgs {RGS}
30:
31: \renewcommand{\textfraction}{0.05}
32: \renewcommand{\topfraction}{0.9}
33: \renewcommand{\bottomfraction}{0.5}
34: \renewcommand{\floatpagefraction}{0.95}
35: \renewcommand\dbltopfraction{0.9}
36: \renewcommand\dblfloatpagefraction{0.95}
37:
38: \begin{document}
39:
40: %\keywords{\LaTeX; ESA; X-rays}
41:
42: \maketitle
43:
44: \begin{abstract}
45: We present and discuss high-resolution grating spectra of the
46: quasar \pgt\ obtained over three years. Based on an early
47: observation from 2001, we find an outflow component of about 3000
48: \kms\ in contrast with the much higher velocity of about 24000
49: \kms\ reported earlier for this source, and based on the same data
50: set. Subsequent grating spectra obtained for \pgt\ are consistent
51: with the first observation in the broad-band sense, but not all
52: narrow features used to identify the outflow are reproduced. We
53: demonstrate that the poor S/N and time variability seen during all
54: existing observations of \pgt\ make any claims about the outflow
55: precariously inconclusive.
56: \end{abstract}
57:
58: \section{Introduction}
59:
60: Typical mass outflow velocities of a few hundreds to a few thousands
61: \kms\ have been measured by now in numerous Active Galactic Nuclei
62: (AGNs; Crenshaw et al. 2003 and references therein). Recent studies
63: of the X-ray spectra of certain quasars have led to claims of much
64: higher outflow velocities reaching a significant fraction of the
65: speed of light, e.g., APM\,08279+5255 --- Chartas et al. (2002)
66: claim speeds of $\sim0.2c$ and $\sim0.4c$\,\footnote{Though Hasinger
67: et al. (2002) using a different instrument prefer a more conservative
68: interpretation by which the X-ray wind is much slower and consistent
69: with the well known, UV broad absorption line wind of that source,
70: outflowing at velocities of up to 12000 \kms .}, PG\,1115+080 ---
71: Chartas et al. (2003) find two X-ray absorption systems with outflow
72: velocities of $\sim0.10c$ and $\sim0.34c$. These measurements,
73: however, were carried out using spectra obtained with CCD cameras
74: and hence at moderate spectral resolving powers of $R\sim 50$. Using
75: the \xmm\ reflection gratings ($R$ up to 500), high resolution
76: X-ray spectra for several quasars have been obtained. For \pgt\
77: Pounds et~al. (2003a, 2005) find a rich, well resolved spectrum featuring
78: absorption lines of several ions, which they interpret as due to an
79: outflow of $\sim$~24000~\kms. A similar interpretation was applied
80: to a similar observation of PG\,0844+349, where Pounds et al. (2003b)
81: report even higher velocities reaching $\sim$~60000~\kms. In NGC\,4051,
82: Pounds et al. (2004) find a single absorption line at $\sim$~7.1~keV,
83: which they suggest may be {Fe}\,{\sc xxvi} Ly$\alpha$ at an
84: outflow velocity of $\sim$~ 6500~\kms, or the He$\alpha$ resonance
85: absorption line of {Fe}\,{xxv} in which case the outflow velocity
86: is $\sim$~16500~\kms. Yet another ultra-high-velocity (UHV, i.e.,
87: sub-$c$) wind of 50000~\kms\ was reported by Reeves et al. (2003)
88: for PDS~456. In all of these sources, the inferred hydrogen column
89: densities through the wind is of the order of 10$^{23}$~\cmii, which
90: is about an order of magnitude higher than the typical values measured
91: for the nearby Seyfert sources.
92:
93: If indeed UHV outflows are common to bright quasars, this could have
94: far reaching implications on our understanding of AGN winds and AGNs
95: in general. For instance, if these winds carry a significant amount
96: of mass as the high column densities may suggest, they would alter
97: our estimates of the metal enrichment of the intergalactic medium by
98: quasars. It remains to be shown theoretically what mechanism (e.g.,
99: radiation pressure) can drive these intense winds. Since the amount
100: of mass in the wind is not well constrained, it is still unclear what
101: effect it may have on the energy budget of the AGN. King \& Pounds
102: (2003) note that UHV winds have been found mostly for AGNs accreting
103: near their Eddington limit. They provide a theory by which the UHV
104: outflows are optically thick producing an effective photosphere,
105: which is also responsible for the UV blackbody and soft X-ray (excess)
106: continuum emission observed for these sources.
107:
108: \section{\pgt\ \ --- \ first observation \ --- \ second view}
109:
110: \begin{figure}
111: \centering
112: \centerline{\includegraphics[width=7.5cm]{pnmodel_all_thaw.eps}}
113: \caption{The EPIC-pn data and a simple fitted model, which is discussed
114: in \S~\ref{epicpn}. The upper panel shows the model folded through the
115: instrument response and compared with the data. Bottom panel shows
116: the unfolded model.}
117: \label{pnmodel}
118: \end{figure}
119:
120: \pgt\ was observed with {\it XMM-Newton} during 2001 June 15 for
121: about 55 ks. We retrieved the data for this observation from the
122: {\it XMM-Newton} archive and reduced them using the Science Analysis
123: System (SAS v5.3.0) in the standard processing chains as described in
124: the data analysis threads and the ABC Guide to {\it XMM-Newton} Data
125: Analysis. Overall, our data reduction results agree well with that of
126: Pounds et al. (2003a, 2005) except for a few minor features which appear
127: to be slightly different between the two reductions. We attribute
128: these discrepancies to the different binning methods used and to the
129: averaging of RGS1 and RGS2 in this work (see below). The results
130: described in this section are presented in details in Kaspi \& Behar
131: (2006).
132:
133: \subsection{EPIC-pn}
134: \label{epicpn}
135:
136: For the EPIC-pn data we first fitted the (line-free) rest-frame 2--5
137: keV energy range with a simple power law. The best fitted power law
138: has a photon index of $\Gamma = 1.55\pm 0.05$ and a normalization
139: of $(6.6\pm 0.4)\times10^{-4}$ ph\,cm$^{-2}$\,s$^{-1}$\,keV$^{-1}$
140: and gives $\chi^{2}_{\nu}=0.74$ for 487 degrees-of-freedom
141: (d.o.f.). Extrapolating this power law up to a rest-frame energy of
142: 11~keV, we find a flux excess above the power law at around 6.4~keV,
143: which is indicative of an iron K$\alpha$ line, and a flux deficit
144: below the power law at energies above 7~keV. We add to the model
145: a Gaussian emission line to account for the Fe\,K$\alpha$ line and
146: a photoelectric absorption edge to account for the deficit. Fitting
147: for all parameters simultaneously, we find the best-fit Gaussian line
148: center is at $6.04\pm 0.04$~keV (or $6.53\pm 0.05$~keV in the rest
149: frame) and a line width ($\sigma$) of $0.096\pm 0.067$~keV. The total
150: flux in the line is $(2.9\pm 1.4)\times 10^{-6}$ \phcms. For the edge,
151: we find a threshold energy of $6.72\pm 0.10$~keV, which is translated
152: to a rest frame energy of $7.27\pm 0.11$~keV. The optical depth at the
153: edge is $\tau=0.56\pm 0.10$. The power law model with the Gaussian
154: line and the absorption edge gives a $\chi^{2}_{\nu}=0.983$ for 613
155: d.o.f. This model is plotted in Figure~\ref{pnmodel} where we show the
156: model, both folded through the instrument and fluxed (i.e., unfolded).
157: We stress that this edge does not necessarily contradict the presence
158: of the line detected by Pounds et al. (2003a, 2005), since K$\alpha$ edges
159: have lines right next to them.
160:
161: We also observe the lines at 2.68 keV and 1.47 keV claimed by Pounds
162: et al. (2003a, 2005) to be from S and Mg, only we identify them as different
163: lines at much lower velocities. The 2.68 keV line is identified here
164: as {S}\,{\sc xv}\,He$\beta$ and the 1.47 keV line is identified as
165: {Mg}\,{\sc xi}\,He$\beta$.
166:
167: \subsection{RGS}
168:
169: \begin{figure*}
170: \centering
171: \includegraphics[width=16cm]{pgtrgs3pan.eps}
172: \caption{Combined RGS1 and RGS2 spectrum of \pgt\ binned to $\sim 0.04$
173: \AA. The spectrum has been corrected for Galactic absorption and for
174: the redshift of the source. The rest-frame positions of lines from
175: H-like and He-like ions of N, O, Ne, and Mg are marked
176: above the spectrum. The lines from the lower ionization states of
177: O and Mg, and the L-shell lines of Fe are not marked. Gaps in the
178: spectrum are due to chip gaps and have zero flux. The model
179: is marked as the solid red curve.}
180: \label{pgtrgs}
181: \end{figure*}
182:
183: The RGS1 and RGS2 were operated in the standard spectroscopy mode
184: resulting in an exposure time of $\sim 52$ ks. The spectra were
185: extracted into uniform bins of $\sim$\,0.04 \AA\ (which is about
186: the RGS resolution and is 4 times the default bin width) in order to
187: increase the signal-to-noise ratio (S/N). For the purpose of modeling
188: narrow absorption lines, this rebinning method is better than the
189: method used by Pounds et al. (2003a) of rebinning the spectrum to a
190: minimum of 20 counts per bin, which distorts the spectrum especially
191: around low-count-rate absorption troughs. To flux calibrate the
192: RGS spectra we divided the count spectrum of each instrument by
193: its exposure time and its effective area at each wavelength. Each
194: flux-calibrated spectrum was corrected for Galactic absorption and the
195: two spectra were combined into an error-weighted mean. At wavelengths
196: where the RGS2 bins did not match exactly the wavelength of the
197: RGS1 bins, we interpolated the RGS2 data to enable the averaging.
198: The sky-subtracted combined RGS spectrum has in total $\sim 8900$
199: counts and its S/N ranges from $\sim 2$ around 8\,\AA\ to $\sim 5$
200: around 18\,\AA\ with an average of 3. Statistics in the second
201: order of refraction are insufficient, hence we did not include it in
202: our analysis.
203:
204: The combined RGS spectrum (RGS1 and RGS2) of \pgt\ is presented in
205: Figure~\ref{pgtrgs}. Numerous absorption lines and several emission
206: lines are detected. We identify K-shell lines of C, N, O and Mg and
207: L-shell lines of O, Mg, Si, Ar, and Fe. The absorption line widths
208: are consistent with the RGS resolution, and with the present S/N we
209: are not able to resolve the intrinsic velocity widths. In emission,
210: we identify significantly broadened lines of {N}\,{\sc vii} Ly$\alpha$,
211: {O}\,{\sc viii} Ly$\alpha$, the forbidden line of {O}\,{\sc vii} and
212: its He$\alpha$ resonance line, the forbidden line of {Ne}\,{\sc ix},
213: and the {Mg}\,{\sc xi} He$\alpha$ resonance line, all in the rest
214: frame of the source with no velocity shift.
215:
216:
217: In order to quantitatively explore the emission and absorption lines,
218: we have constructed a model for the entire RGS spectrum. The present
219: method is an ion-by-ion fit to the data similar to the approach used
220: in Sako et al. (2001) and in Behar et~al. (2003). We first use the
221: continuum measured from the EPIC-pn data, but renormalized to the
222: RGS flux level. This continuum is then absorbed using the full set
223: of lines for each individual ion. Our absorption model includes
224: the first 10 resonance lines of H- and He- like ions of C, N, O,
225: Ne, and Mg as well as edges for these ions. The model also includes
226: our own calculation for the L-shell absorption lines of Fe (Behar et
227: al. 2001) as well as of Si, S, and Ar corrected according to laboratory
228: measurements (Lepson et al. 2003, 2005). Finally, we include inner-shell
229: K$\alpha$ absorption lines of O and Mg (Behar \& Netzer 2002), which
230: we detect in the spectrum. The absorbed spectrum is complemented by
231: the emission lines mentioned above, which are observed in the RGS
232: spectrum.
233:
234: By experimenting with the absorption line parameters, we find that
235: the observed lines are all blueshifted by about 3000~\kms\ with an
236: uncertainty of 500~\kms . In the model we used a turbulence velocity
237: of 1000~\kms\ to broaden the absorption lines. This width includes
238: the instrumental broadening, which as noted above, we could not
239: separate from the intrinsic broadening. Since the lines appear to
240: be saturated, but no line goes to zero intensity in the trough, we
241: obtain the best fit by assuming a covering factor of 0.7 for the X-ray
242: continuum source. The best-fit column densities that we find for the
243: different ions are consistent with a hydrogen column density of about
244: $10^{21}$--$10^{22}$ \cmii . The emission lines are modeled using
245: Gaussians with uniform widths of $\sigma = 2500$ \kms\ (resolved, but
246: again, including the instrumental broadening), with no velocity shift,
247: and assumed to be unabsorbed. These lines at FWHM~$\simeq$ $6000\pm
248: 1200$~\kms\ are even broader than those observed from the broad line
249: region in the visible band ($\sim 2000$ \kms ; Kaspi et al. 2000). The
250: entire best-fit spectrum is shown in Figure~\ref{pgtrgs} (red curve).
251: The spectrum beyond 25~\AA\ is particularly challenging as it comprises
252: many unresolved lines from L-shell ions of Si, S, and Ar while the
253: RGS effective area drops rapidly. Several predicted lines may be
254: observed here (e.g., {Ar}\,{\sc xiii} - 28.92~\AA, {Si}\,{\sc xii} -
255: 30.71~\AA, {Ar}\,{\sc xii} - 31.06~\AA, {S}\,{\sc xiii} - 31.93~\AA;
256: these wavelengths include the 3000 \kms\ shift). We are still unable to
257: explain several features seen in the data, e.g., around 8.5 \AA, 10.4
258: \AA, or 29.8 \AA , but the model gives a good fit to the data overall.
259:
260:
261: \subsection{Conclusions - First Observation}
262:
263:
264: We have provided a self consistent model to the ionized outflow
265: of \pgt\ revealing an outflow velocity of approximately 3000~\kms.
266: Our model reproduces many absorption lines in the RGS band, although
267: the S/N of the present data set is rather poor and some of the noise
268: might be confused with absorption lines.
269:
270: The present approach is distinct from the commonly used global
271: fitting methods and also from the line-by-line approach used by
272: Pounds et al. (2003a). It allows for a physically consistent
273: fit to the spectrum and is particularly appropriate for a
274: broad-ionization-distribution absorber as observed here for \pgt .
275:
276: The present model also features several broad (FWHM = 6000~\kms)
277: emission lines, which are observed directly in the data.
278:
279: A broad and relatively flat ionization distribution is found
280: throughout the X-ray outflow consistent with a hydrogen column density
281: of roughly $10^{21}$--$10^{22}$ \cmii. This is reminiscent of the outflow
282: parameters measured in other well studied Seyfert galaxies.
283:
284: We also detect Fe-K absorption, which was identified by Pounds
285: et~al. (2003a, 2005) as a strongly blueshifted {Fe}\,{\sc xxvi} absorption
286: line. We find that most of the Fe-K opacity can alternatively be
287: attributed to several consecutive, low charge states of Fe, although
288: it can not be assessed whether the absorber is co-moving with the
289: outflow or not. Future missions with microcalorimeter spectrometers
290: on board might be able to address this interesting question.
291:
292:
293: \section{A second Observation of \pgt}
294:
295: \begin{figure*}
296: \centering
297: \includegraphics[width=16.2cm]{2rgs.eps}
298: \caption{Two RGS spectra of \pgt\ obtained three years apart. The
299: first observation (in black) was carried out on 2001 June 15 and
300: is also presented in Figure 1. The second observation (in red)
301: took place on 2004 June 21.} \label{2rgs}
302: \end{figure*}
303:
304: A second \xmm\ observation of \pgt\ for $\sim 50$\,ks was carried
305: out on 2004 June 21, three years after the first observation of 2001
306: June 15 which is described above. We have retrieved the data of this
307: second observation from the \xmm\ archive and reduced it in exactly the
308: same way described above for the first observation. The data of the
309: second RGS observation are plotted (red line) in Figure~\ref{2rgs}
310: over the first observation (black line). The broad-band spectra
311: of the two observations are generally consistent. However, not all
312: narrow features are consistently reproduced. The total flux in the
313: RGS band is the same in the two observations, though the continuum
314: slope in the second observation the spectrum is somewhat harder.
315:
316: When inspecting the detailed narrow features in the spectrum some
317: have changed while others remain the same. For example, the second
318: observations shows features which appear to be emission lines around
319: 8 \AA\, where the first observation had absorption lines. Also,
320: around 15 \AA\ the absorption lines seem to have disappeared in the
321: second observation. Conversely, some features are the same in the two
322: spectra, for example, the emission O\,{\sc vii} triplet around 22 \AA\
323: and the Ne\,{\sc ix} triplet around 13.5 \AA. From Figure~3, it can
324: be seen that due to the poor S/N in both spectra, it is extremely hard
325: to determine whether the differences between the two spectra are real,
326: or a mere result of the data's poor S/N.
327:
328: \section{Simultaneous observations of \xmm\ and {\it Chandra}}
329:
330: Simultaneously with the 2004 June 21 \xmm\ observation, \pgt\ was
331: also observed with the Low Energy Transmission Grating (LETG) on
332: board the {\it Chandra} X-ray observatory. The LETG observation of
333: $\sim 45$~ks has made use of the ACIS CCDs as the detector. We
334: retrieved the data of this observation from the {\it Chandra
335: archive} and reduced it using CIAO 3.2.1 and CALDB version 3.01,
336: according to the updated CIAO threads. We have combined the +1
337: and -1 orders of the LETG spectrum using a weighted mean and the
338: combined spectrum is represented in Figure~\ref{rgs_letg} by a
339: black line. The simultaneous RGS observation (the red data in
340: Figure~\ref{2rgs}) is shown in red in Figure~\ref{rgs_letg}.
341:
342: Although the simultaneous data from the two X-ray observatories
343: are consistent overall, they differ in many details. For example,
344: the RGS data between 7 to 9 \AA\ show several emission-like
345: features which the LETG data do not. Also, around 16.4 \AA\ the
346: LETG data show absorption-like features which are not present in
347: the RGS data. These differences are consistent to within about
348: 3$\sigma$ and are probably a result of the poor S/N of the
349: observations
350:
351: After the first $\sim$\,45\,ks LETG observation of \pgt\ on 2004
352: June 21, which is described above, there were two more observations
353: in consecutive orbits of the {\it Chandra} observatory. The second
354: $\sim 45$\,ks observation took place on 2004 June 23 and the third
355: observation was on 2004 June 25. These data are not presented here, but
356: their spectra is in overall agreement with the first LETG observation,
357: {\it except} that the flux level in the last two observations was
358: twice that of the first observation, i.e., during a period of $\sim
359: 2$ days the flux level doubled. This is somewhat unexpected for
360: a source that had retained its flux level of three years earlier
361: (see Figure~\ref{2rgs}). Besides the change in flux between the three
362: LETG observations, there is also a change in the absorption features
363: seen between the first observation and the other two. Some of these
364: features have disappeared between the first low-flux observation and
365: the high-flux observations taken 2 days later, while other features
366: seems to appear. The absorption features interpreted by Reeves et
367: al. (2005) as evidence for sub-c gravitational infall are seen only
368: in the second observation and not in the first or third ones. The
369: fact that, again, absorption features are not reproduced in different
370: spectra is rather confusing. If these lines are statistically
371: significant (see Reeves et al. 2005) then they represent a transient flow.
372:
373: \section{Summary and Conclusions}
374:
375: We claim in Kaspi \& Behar (2006) that an outflowing absorber
376: at a velocity of 3000 \kms\ fits the first (2001) RGS data of
377: \pgt\ better than a 24000 \kms\ model. Admittedly though, the poor
378: S/N of those data can tolerate more than one interpretation.
379:
380: A second RGS observation taken three years after the first
381: observation shows general consistency with the first observation,
382: but differs in important details of the absorption lines relevant
383: to the outflow. Some features that appear in the first RGS
384: observation disappear in the second one and vice versa. This could
385: be a result of either short-time variability of the absorber
386: (almost impossible to prove or refute) or the poor S/N of the
387: data. Even more confusing is the fact that {\it simultaneous}
388: observations of \pgt\ with RGS and LETG produce spectra that are
389: partially incompatible in their absorption lines. This
390: significantly reduces our confidence in the existence of the
391: absorption lines and even more so in their identification. The
392: poor S/N of the data calls for extra caution and careful modeling.
393:
394: The three LETG observations indicate that the continuum source
395: changes on a timescale of days. If the discrete features seen in
396: these spectra are real, they too vary on short timescales. With
397: the loss of the high-resolution X-ray spectrometer (XRS) on board
398: {\it Astro-E2}, a very long observation of a good, bright UHV-wind
399: source with {\it Chandra} or \xmm\ gratings remains as the most
400: viable approach toward testing what we feel is still a putative
401: phenomenon of high velocity outflow.
402:
403: \section*{Acknowledgments}
404:
405: This research was supported by the Israel Science Foundation (grant no.
406: 28/03), and by a Zeff fellowship to S.K.
407:
408: \begin{figure*}
409: \centering
410: \includegraphics[width=17cm]{rgs_letg.eps}
411: \caption{ \pgt\ spectrum taken by the RGS on board \xmm\ (red line)
412: on 2004 June 21 shown together with a spectrum taken simultaneously
413: by the LETG on board {\it Chandra} (black line). Although the data are
414: taken at the same time, discrepancies between the spectra are evident,
415: probably as a result of the poor S/N of the data.}
416: \label{rgs_letg}
417: \end{figure*}
418:
419:
420: \begin{thebibliography}{}
421:
422: \bibitem[Behar, Cottam, \& Kahn (2001)]{2001ApJ...548..966B}
423: Behar,~E., Cottam,~J.~C., \& Kahn,~S.~M. 2001, \apj, 548, 966
424: % Capella
425:
426: \bibitem[Behar & Netzer (2002)]{2002ApJ...570..165B}
427: Behar, E., \& Netzer, H. 2002, \apj, 570, 165
428:
429: \bibitem[Behar et~al. (2003)]{2003ApJ...598..232B}
430: Behar,~E., Rasmussen,~A.~P., Blustin,~A.~J., Sako,~M.,
431: Kahn,~S.~M., Kaastra,~J.~S., Branduardi-Raymont,~G., \&
432: Steenbrugge,~K.C. 2003, \apj, 598, 232
433: % NGC3783
434:
435: \bibitem[Chartas, Brandt, Gallagher, \& Garmire(2002)]{2002ApJ...579..169C}
436: Chartas, G., Brandt, W.~N., Gallagher, S.~C., \& Garmire, G.~P.\ 2002,
437: \apj, 579, 169
438: % APM08279+5255
439:
440: \bibitem[Chartas, Brandt, \& Gallagher(2003)]{2003ApJ...595...85C} Chartas,
441: G., Brandt, W.~N., \& Gallagher, S.~C.\ 2003, \apj, 595, 85
442: % PG1115+080
443:
444: \bibitem[Crenshaw, Kraemer, \& George(2003)]{2003ARA&A..41..117C} Crenshaw,
445: D.~M., Kraemer, S.~B., \& George, I.~M.\ 2003, \araa, 41, 117
446:
447: \bibitem[Hasinger, Schartel, \& Komossa(2002)]{2002ApJ...573L..77H}
448: Hasinger, G., Schartel, N., \& Komossa, S.\ 2002, \apjl, 573, L77
449: % APM08279+5255
450:
451: \bibitem[Kaspi \& Behar (2006)]{2006ApJ}Kaspi, S., \& Behar, E. 2006,
452: ApJ, 636, in press
453:
454: \bibitem[Kaspi et al.(2000)]{2000ApJ...533..631K} Kaspi, S., Smith, P.~S.,
455: Netzer, H., Maoz, D., Jannuzi, B.~T., \& Giveon, U.\ 2000, \apj, 533, 631
456:
457: \bibitem[King \& Pounds(2003)]{2003MNRAS..345...657K}
458: King,~ A.~R. \& Pounds,~K.~A. 2003, \mnras, 345, 657
459:
460: \bibitem[Lepson et al. (2003)]{jaan03} Lepson, J.~K., Beiersdorfer, P.,
461: Behar,~E., \& Kahn,~S.~M. 2003, \apj, 590, 604
462:
463: \bibitem[Lepson et al.(2005)]{2005ApJ...625.1045L} Lepson, J.~K.,
464: Beiersdorfer, P., Behar, E., \& Kahn, S.~M.\ 2005, \apj, 625, 1045
465:
466: \bibitem[Pounds et al.(2003a)]{2003MNRAS.345..705P} Pounds, K.~A., Reeves,
467: J.~N., King, A.~R., Page, K.~L., O'Brien, P.~T., \& Turner, M.~J.~L.\ 2003a,
468: \mnras, 345, 705
469: %PG1211+143
470:
471: \bibitem[Pounds, King, Page, \& O'Brien(2003b)]{2003MNRAS.346.1025P} Pounds,
472: K.~A., King, A.~R., Page, K.~L., \& O'Brien, P.~T.\ 2003b, \mnras, 346, 1025
473: % PG0844+349
474:
475: \bibitem[Pounds, Reeves, King, \& Page(2004)]{2004MNRAS.350...10P} Pounds,
476: K.~A., Reeves, J.~N., King, A.~R., \& Page, K.~L.\ 2004, \mnras, 350, 10
477: % NGC4051
478:
479: \bibitem[Pounds et al.(2005)]{2005P} Pounds, K.~A., Reeves, J.~N., King, A.~R.,
480: Page, K.~L., O'Brien, P.~T., \& Turner, M.~J.~L.\ 2005, \mnras, 356, 1599
481: %PG1211+143 Erratum
482:
483: \bibitem[Reeves, O'Brien, \& Ward,(2003)]{2003ApJ...593L..65R}
484: Reeves, J.~N., O'Brien,~P.~T., \& Ward, M.~J. 2003, \apj, 593, 65
485: % PDS456
486:
487: \bibitem[Reeves, et~al.(2005)]{reeves05}
488: Reeves, J.~N., Pounds,K., Uttley, P., Kraemer, S., Mushotzky, R., Yaqoob, T.,
489: George, I.~M., and Turner, T.~J. 2005, ApJL, 633, L81
490: % astro-ph/0509280
491:
492: \bibitem[Sako et al. (2001)]{sako01} Sako, M., et~al. 2001, \aap, 365, L168
493:
494: \end{thebibliography}
495:
496: \end{document}
497: