1: \documentclass[12pt,preprint]{aastex}
2: %\usepackage{epsfig}
3:
4: \shortauthors{Mattson, Weaver \& Reynolds}
5: \shorttitle{\RXTE\ Seyfert 1 Observations}
6:
7: \begin{document}
8:
9: \newcommand{\RXTE}{\textit{RXTE}}
10: \newcommand{\ASCA}{\textit{ASCA}}
11: \newcommand{\suzaku}{\textit{Suzaku}}
12:
13: \title{Possible X-ray diagnostic for jet/disk dominance in Type 1 AGN}
14:
15: \author{Barbara J. Mattson\altaffilmark{\star}, Kimberly A.
16: Weaver}\affil{NASA/Goddard Space Flight Center, Astrophysics Science
17: Division, Greenbelt, MD, 20771}
18:
19: \and
20: \author{Christopher S. Reynolds}\affil{Department of Astronomy,
21: University of Maryland, College Park, MD, 20742}
22:
23: \altaffiltext{\altaffilmark{\star}}{also Department of Astronomy,
24: University of Maryland, College Park, MD and Adnet Systems, Inc.,
25: Rockville, MD}
26:
27: \begin{abstract}
28: Using \textit{Rossi X-ray Timing Explorer} Seyfert 1 and 1.2 data
29: spanning 9 years, we study correlations between X-ray spectral features.
30: The sample consists of 350 time-resolved spectra from 12 Seyfert 1 and
31: 1.2 galaxies. Each spectrum is fitted to a model with an intrinsic
32: powerlaw X-ray spectrum produced close to the central black hole that is
33: reprocessed and absorbed by material around the black hole. To test the
34: robustness of our results, we performed Monte Carlo simulations of the
35: spectral sample. We find a complex relationship between the iron line
36: equivalent width ($EW$) and the underlying power law index ($\Gamma$).
37: The data reveal a correlation between $\Gamma$ and $EW$ which turns over
38: at $\Gamma \lesssim 2$, but finds a weak anti-correlation for steeper
39: photon indices. We propose that this relationship is driven by dilution
40: of a disk spectrum (which includes the narrow iron line) by a beamed jet
41: component and, hence, could be used as a diagnostic of jet-dominance.
42: In addition, our sample shows a strong correlation between $R$ and
43: $\Gamma$, but we find that it is likely the result of modeling
44: degeneracies. We also see the X-ray Baldwin effect (an anti-correlation
45: between the 2-10 keV X-ray luminosity and $EW$) for the sample as a
46: whole, but not for the individual galaxies and galaxy types.
47: \end{abstract}
48:
49: \keywords{galaxies: Seyfert, X-rays: galaxies}
50:
51:
52: \section{Introduction} \label{section:intro}
53:
54: Time-resolved X-ray spectroscopy studies of active galactic nuclei (AGN)
55: offer the opportunity to investigate emission regions near the central
56: black hole. In fact, X-ray spectroscopy offers the clearest view of
57: processes occurring very close to the black hole itself, probing matter
58: to its final plunge into the black hole. Armed with such information,
59: we can unlock the structure of the innermost regions of AGN.
60:
61: Typical X-ray spectra of AGN show an underlying powerlaw produced near
62: the central black hole with signatures of reprocessed photons often
63: present. These reprocessed photons show up as an Fe K$\alpha$ line at
64: $\sim$6.4 keV and a ``reflection hump'' which starts to dominate near 10
65: keV and is produced by the combined effects of photoelectric absorption
66: and Compton downscattering in optically-thin cold matter irradiated by
67: the hard X-ray continuum. The Fe K$\alpha$ line has been observed in
68: both type 1 (unabsorbed) and type 2 (absorbed) Seyfert galaxies. It has
69: been attributed to either the broad line region, the accretion disk, the
70: molecular torus of unification models \citep{antonucci93}, or some
71: combination of these. Signatures of reflection have also been observed
72: in both Seyfert 1 and 2 galaxies.
73:
74: If the unification models are correct, we should see similar spectral
75: correlations between Seyfert 1 and 2 galaxies, with any differences
76: easily attributable to our viewing angle. Regardless of the accuracy of
77: the reflection models, we expect changes in the underlying continuum to
78: drive changes in the reprocessing features. However, results from X-ray
79: spectral studies of AGN have so far produced puzzling results. Samples
80: of Seyfert 1 observations from \ASCA\ \citep{weaver01} and \textit{Rossi
81: X-ray Timing Explorer} \citep{markowitz03} have shown no obvious
82: relationship between changes in the continuum and iron line. Several
83: galaxies have shown an anticorrelation between reflection and/or iron
84: line equivalent width and the source flux; e.g. NGC 5548
85: \citep{chiang00}, MCG $-$6-30-15 \citep{papadakis02}, NGC 4051
86: \citep{papadakis02, wang99}, NGC 5506 \citep{papadakis02, lamer00}.
87: Recent data from \suzaku\ on MCG $-$6-30-15, on the other hand, show
88: that the iron line and reflection remain relatively constant while the
89: powerlaw is highly variable \citep{miniutti06}. \citet{zdziarski99}
90: found that Seyfert galaxies and X-ray binaries show a correlation
91: between the continuum slope and reflection fraction, so those with soft
92: intrinsic spectra show stronger reflection than those with hard spectra.
93: However, other studies have found either a shallower relationship than
94: Zdziarski et al. \citep{perola02} or an anticorrelation
95: \citep{papadakis02, lamer00}.
96:
97: Here we present the first results of a larger study of the X-ray
98: spectral properties of Seyfert galaxies observed by the \textit{Rossi
99: X-ray Timing Explorer} (\RXTE). Our full study consists of observations
100: of 30 galaxies. In this letter, we focus on the spectral results from
101: the subset of 12 Seyfert 1 and 1.2 galaxies. In
102: \S~\ref{section:analysis} we present our method of data analysis,
103: including our sample selection criteria (\S~\ref{subsection:sample}), a
104: description of our data pipeline (\S~\ref{subsection:pipeline}), and
105: results of our spectral analysis (\S~\ref{subsection:results}). We
106: discuss the implications of our results in \S~\ref{section:discussion}
107: and detail our conclusions in \S~\ref{section:conclusions}.
108:
109:
110:
111: \section{Data Analysis} \label{section:analysis}
112:
113: \subsection{The Sample} \label{subsection:sample}
114:
115: The \RXTE\ public archive\footnote{Hosted by the High Energy
116: Astrophysics Science Archive and Research Center (HEASARC;
117: \url{http://heasarc.gsfc.nasa.gov/})} represents one of the largest
118: collections of X-ray data for AGN, with pointed observations of over 100
119: AGN spanning 10 years. The \RXTE\ bandpass allows the study of
120: absorption and iron line properties of AGN spectra, as well as a glimpse
121: at the Compton reflection hump. We use data from the \RXTE\
122: proportional counter array (PCA), which is sensitive to energies from 2
123: to 60 keV and consists of five Proportional Counter Units (PCUs). Most
124: of the sources in our sample do not show significant counts in the
125: \RXTE\ Hard Energy X-ray Timing Experiment (HEXTE), so we do not include
126: HEXTE data in this study.
127:
128: To focus this study, we choose only Seyfert galaxies for which the
129: \RXTE\ public archive contained a minimum of two pointings separated by
130: at least two weeks. We further required the total observed time be
131: $>40$ ks. These selection criteria led to a sample of 40 Seyfert
132: galaxies. For the analysis presented here, we examine the 18 Seyfert 1
133: and 1.2 galaxies. Six galaxies were eliminated after they were put
134: through our data pipeline (see \S~\ref{subsection:pipeline} for more),
135: so the final sample presented here consists of 12 galaxies, listed in
136: Table~\ref{table:sources}. Because the data come from the public
137: archive, the sample is not uniform from galaxy to galaxy or even from
138: observation to observation; however, we use the Standard 2 data, which
139: provides a standard data mode for these diverse observations.
140:
141:
142: \subsection{Data Pipeline} \label{subsection:pipeline}
143:
144: To ensure consistent data reduction of the large volume of data, we
145: developed a data pipeline. The Standard 2 data for each observation was
146: reduced using a combination of FTOOLs and the Python$^{\circledR}$ scripting
147: language. The pipeline produces time-resolved spectra, each with a
148: minimum of 125,000 net photons, which are extracted using standard PCA
149: selection criteria and background models \citep{jahoda06}. Sources
150: which did not have sufficient net photons for even one spectrum were
151: eliminated from the final sample (Table~\ref{table:sources} shows the
152: final sample with the 6 eliminated sources listed in the table
153: notes). Each spectrum includes 1\% systematic errors. We are confident
154: in the instrument response and background models up to energies of
155: $\sim$25 keV, so we ignore channels with higher energies.
156:
157:
158: \subsection{Spectral Fitting and Results} \label{subsection:results}
159:
160: The data pipeline produced 350 spectra for the 12 galaxies in our
161: sample. Each spectrum was fitted from 3 to 25 keV with an absorbed
162: Compton reflection model plus a Gaussian iron line. In \texttt{xspec},
163: the PEXRAV \citep{magdziarz95} model simulates the effects of an
164: exponentially cut-off powerlaw reflected by neutral matter and has seven
165: model parameters: photon index of the intrinsic underlying power-law
166: ($\Gamma$), the cutoff energy of the power law in keV ($E_c$), the
167: relative amount of reflection ($R$), the redshift ($z$), the abundance
168: of heavy elements in solar units ($Z$), the disk inclination angle
169: ($i$), and the photon flux of the power law at 1 keV in the observer's
170: frame ($A$). The relative amount of reflection is normalized to 1 for
171: the case of an isotropic source above a disk of neutral material
172: ($\Omega=2\pi$). Adding a Gaussian line (energy in keV ($E_{Fe}$),
173: physical width ($\sigma$) in keV, and normalization in units of photons
174: cm$^{-2}$ s$^{-1}$) and an absorbing column ($N_H$, in cm$^{-2}$) yields
175: a total of 12 parameters.
176:
177: We fixed the following values in PEXRAV: $E_c$ = 500 keV, $Z$ = 1.0, and
178: $\cos i$ = 0.95. This inclination represents an almost face-on disk;
179: however, since we are seeking trends in the spectral parameters, rather
180: than absolute values, the precise value is not important to this study.
181: In addition, $z$ is fixed at the appropriate value from the NASA
182: Extragalactic Database for each
183: galaxy\footnote{\url{http://nedwww.ipac.caltech.edu/}}. After fitting
184: all spectra to this model, we derived the mean Gaussian width for each
185: source (Table~\ref{table:sources}), then held $\sigma$ fixed for a
186: second fit to the model. Our final model has free parameters: $\Gamma$,
187: $R$, $A$, $E_{Fe}$, iron line normalization and $N_H$. To prevent
188: \texttt{xspec} from pursuing unphysical values of the parameters, we set
189: the following hard limits: $0 \leq \Gamma \leq 5$, $0 \leq R \leq 5$, $
190: 5.5 \leq E_{Fe} \leq 7.5$ keV, and $0 \leq \sigma \leq 1.5$ keV (for the
191: free-$\sigma$ fits).
192:
193: Looking at the iron line equivalent width ($EW$) and $\Gamma$, we find a
194: complex relationship with a ``hump'' peaking near $\Gamma \sim 2.0$
195: (Figure~\ref{fig1}a). The $EW$-$\Gamma$ plot shows a correlation for
196: $\Gamma \lesssim 2.0$ and an anti-correlation for $\Gamma \gtrsim 2.0$,
197: with a peak near $\Gamma \sim 2.0$ with $EW \sim 250$ eV. We also find
198: a strong correlation between $R$ and $\Gamma$ (Figure~\ref{fig2}a), with
199: a best-fit line of $R = -0.87 + 0.54 \: \Gamma$ ($\chi^2 = 506/349 =
200: 1.46$).
201:
202: We performed a Monte Carlo simulation to determine if our results were
203: an artifact of modeling degeneracies. Each spectrum in the Monte Carlo
204: sample was simulated with $N_H$=$10^{22}$ cm$^{-2}$, $\Gamma$=2.0,
205: $R$=1.0, $E_{Fe}$=6.4 keV, and $\sigma_{Fe}$=0.23 keV. The flux and
206: exposure times were randomly varied for each spectrum. The flux was
207: varied by randomly choosing $A$ from a uniform distribution between
208: 0.004 and 0.06 photons keV$^-1$ cm$^-2$ s$^-1$. The exposure time was
209: randomly generated from a uniform distribution between 300 and 11000s.
210: The ranges for $A$ and the exposure time represent the range of $A$ and
211: exposure for the spectra in the full sample.
212:
213: We generated 200 spectra: 100 simulated using an \RXTE\ Epoch 3
214: response, 50 using an Epoch 4 response, and 50 using an Epoch 5
215: response, roughly corresponding to our \RXTE\ sample. Each spectrum was
216: then fitted to the same model as our full sample. The $R$ over $\Gamma$
217: plot (Figure~\ref{fig2}b) clearly shows a strong correlation with a
218: best-fit line of $R=-7.3+4.1 \: \Gamma$ ($\chi^2 = 28.96/159 =
219: 0.182$), which strongly suggests that the observed $R$-$\Gamma$
220: correlation is a result of modeling degeneracies. The correlation shows
221: a much steeper relationship than the Seyfert 1 data, due to the
222: large number of Seyfert 1 spectra showing $R \sim 0$.
223:
224: $EW$ and $\Gamma$, however, do not suffer the same degeneracies, which
225: is clear from the Monte Carlo results (Figure~\ref{fig1}b). Based
226: on the lack of correlation in our Monte Carlo results, we are confident
227: that the shape of the $EW$-$\Gamma$ plot for the data sample is real.
228:
229: To further examine the $EW$-$\Gamma$ relationship, we reproduced the
230: $EW$-$\Gamma$ plot to show the contribution from each galaxy
231: (Figure~\ref{fig3}). The radio-loud galaxies form the rising leg, with
232: the quasar, 3C 273, anchoring the low $\Gamma$-low $EW$ portion of the
233: plot. The Seyfert 1 (radio quiet) and 1.2 galaxies tend to congregate at
234: the peak and the falling leg of the plot. The one narrow-line Seyfert 1
235: diverges from the main cluster of points.
236:
237: Finally, we examined $EW$ as a function of the intrinsic 2-10 keV X-ray
238: luminosity ($L_x$), using $H_0 = 70$ km s$^{-1}$ Mpc$^{-1}$. We fitted
239: the data for each galaxy, each type, and the sample as a whole to linear
240: and powerlaw models. The data were well-fit for either model. For
241: consistency with other publications, we report here the powerlaw
242: results. For the sample as a whole, we see an anticorrelation, i.e. the
243: X-ray Baldwin effect \citep{iwasawa93}, with $EW \propto
244: L_x^{-0.14\pm0.01}$. When examining galaxy types, however, the
245: anticorrelation does not always hold up (Table~\ref{table:sources}). We
246: find an anticorrelation in the radio loud galaxies and the Seyfert 1.2s,
247: but a marginal correlation for the quasar and radio quiet Seyfert 1s.
248:
249:
250: \section{Discussion} \label{section:discussion}
251:
252: \subsection{$EW$-$\Gamma$ Relationship} \label{section:deiscussion:ewg}
253:
254: The simulations of \citet{george91} for the observed spectrum from an
255: X-ray source illuminating a half-slab showed that the spectra should
256: include a ``Compton hump'' and an iron line. They found that the iron
257: line EW should decrease as the spectrum softens. This is easy to
258: understand, since as the spectrum softens ($\Gamma$ increases), there
259: are fewer photons with energies above the iron photoionization
260: threshold. Our results show that the relationship between $EW$ and
261: $\Gamma$ is not quite so simple. We find a correlation between $EW$ and
262: $\Gamma$ when $\Gamma \lesssim 2$ and an anticorrelation when $\Gamma
263: \gtrsim 2$. Other researchers have found a correlation for Seyfert 1
264: samples \citep{perola02, lubinski01}, but the galaxies in their samples
265: primarily fell in the $\Gamma \lesssim 2$ region. \citet{page04} also
266: find that their data suggest a slight correlation for a sample of radio
267: loud and radio quiet Type 1 AGN.
268:
269: A close examination of our $EW$-$\Gamma$ plot shows that the data for
270: different galaxy types progresses across the plot. The plot is anchored
271: at the low-$\Gamma$, low-$EW$ end by the quasar, 3C 273, in our sample.
272: The rising arm of the plot, $\Gamma \sim 1.5 - 2.0$ and $EW \sim 0 -
273: 300$ eV, is primarily formed by radio loud Seyfert 1 galaxies. The
274: radio-quiet Seyfert 1 galaxies cluster near the $\Gamma \sim 2.0$, $EW
275: \sim 300$ eV peak of the hump, and the radio-quiet Seyfert 1.2 galaxies
276: form the falling arm of the plot for $\Gamma > 2.0$.
277:
278: Physically, the most obvious difference between these sources is the
279: presence or absence of a strong jet. We propose that this relationship
280: is driven by the degree of jet-dominance of the source. The iron line
281: features are associated with the X-ray emission from the disk. Since
282: the disk is essentially isotropic, it will excite an observable iron
283: line from matter out of our line-of-sight. On the other hand, the jet
284: is beamed away from the obvious configurations of matter in the system
285: and, more importantly, is beamed toward us in the quasar and radio-loud
286: sources. Both of these jet-related phenomena reduce the observed
287: equivalent width of any iron line emission associated with the jet
288: continuum.
289:
290: In order for the $\Gamma$ to increase as the jet-dominance decreases,
291: the jet in these sources must have a hard X-ray component, which implies
292: that the radio-loud Seyferts in our sample are to be associated with
293: low-peaked BL Lac objects (LBLs). BL Lac objects show two broad peaks
294: in their spectral energy density plots \citep{giommi94}, with the
295: lower-energy peak due to synchrotron emission and the higher-energy peak
296: due to inverse Compton emission. BL Lacs are divided into two classes,
297: depending on where the peaks occur: high-peaked BL Lacs (HBLs) and LBLs.
298: The X-ray continuum in the HBLs is rather soft, since we are seeing the
299: synchrotron spectrum cutting off in these sources. LBLs, on the other
300: hand, tend to have a harder X-ray continua, since we are observing well
301: into the inverse Compton part of the spectrum \citep{donato05}.
302:
303: We also note that much of the falling arm of the $EW$-$\Gamma$
304: relationship is formed by MCG $-$6-30-15. Recent observations of MCG
305: $-$6-30-15 by Suzaku have shown that the reflection component, including
306: the iron line, remains relatively constant \citet{miniutti06}. We would
307: expect, then, that as $\Gamma$ increases, the $EW$ should decrease,
308: which is exactly what we see in our data.
309:
310:
311: \subsection{$R$-$\Gamma$ Relationship} \label{section:discussion:rg}
312:
313: Significant degeneracies between the photon index, absorbing column, and
314: reflection fraction can easily lead to false conclusions about spectral
315: correlations. These degeneracies occur as these three parameters trade
316: off against each other in the modeling process, an effect that is
317: especially strong in the \RXTE\ bandpass. Our $R$-$\Gamma$ plot shows a
318: strong correlation which is mimicked in our Monte Carlo results. The
319: few points that lie under the main concentration are likely to be
320: outliers, and not indicative of a subclass of galaxy. These points all
321: come from spectra that have been fitted to have $N_H = 0$, and are
322: primarily radio-loud galaxies. We conclude that the observed
323: $R$-$\Gamma$ correlation in our sample cannot be trusted as a real
324: correlation.
325:
326:
327: \subsection{$EW$-$L_x$ Relationship} \label{section:discussion:ewlx}
328:
329: Looking at the $EW$-$L_x$ relationship, we do see the X-ray Baldwin
330: effect for our sample as a whole, with a slighly shallower
331: anticorrelation than reported elsewhere. We find $EW \propto
332: L_x^{-0.14}$, whereas \citet{iwasawa93} and \citet{jiang06} find $EW
333: \propto L_x^{-0.20}$ and \citet{page04} find $EW \propto L_x^{-0.17}$.
334: However, when \citet{jiang06} exclude the radio loud galaxies from their
335: sample, they find $EW \propto L_x^{-0.10}$.
336:
337: We find, though, that when we examine our data on a galaxy-by-galaxy or
338: type-by-type basis, the effect is not consistent from source to source.
339: At this point, we cannot determine if these variations are real or are
340: simply due to the small number of spectra for some of our galaxies and
341: types.
342:
343:
344:
345:
346: \section{Conclusions} \label{section:conclusions}
347:
348: We have examined time-resolved spectra of 12 Seyfert 1 and 1.2 galaxies
349: observed by \RXTE\ over seven years. We find a complex relationship
350: between the iron line equivalent width and the continuum slope, with a
351: correlation for $\Gamma \lesssim 2$ that turns over to an
352: anticorrelation for $\Gamma \gtrsim 2$. We propose that this
353: relationship is a possible diagnostic for jet- versus disk-dominated
354: sources, where jet-dominated sources show a correlation between $EW$ and
355: $\Gamma$, and disk-dominated sources show an anticorrelation. We also
356: see a strong correlation between $\Gamma$ and $R$ which is likely an
357: artifact of modeling degeneracies caused by the interplay of $\Gamma$,
358: $R$, and $n_H$ in the \RXTE\ bandpass. Finally, we observe the X-ray
359: Baldwin effect for the sample as a whole, but not for each galaxy and
360: galaxy type individually.
361:
362:
363:
364: \acknowledgments
365: This research has made use of data obtained from the High Energy
366: Astrophysics Science Archive Research Center (HEASARC), provided by
367: NASA's Goddard Space Flight Center.
368:
369: This research has also made use of the NASA/IPAC Extragalactic Database
370: (NED) which is operated by the Jet Propulsion Laboratory, California
371: Institute of Technology, under contract with the National Aeronautics
372: and Space Administration.
373:
374: CSR gratefully acknowledges support from the National Science
375: Foundation under grants AST0205990 and AST0607428.
376:
377:
378:
379: \begin{thebibliography}{}
380:
381: \bibitem[Antonucci(1993)]{antonucci93}Antonucci, R. 1993, \araa, 31, 473
382:
383: \bibitem[Chiang et al.(2000)]{chiang00}Chiang, J. 2000, \apj, 528, 292
384:
385: \bibitem[Donato, Sambruna, \& Gliozzi(2005)]{donato05}Donato, D.,
386: Sambruna, R. M., \and Gliozzi, M. 2005, \aa, 433, 1163
387:
388: \bibitem[George \& Fabian(1991)]{george91}George, I. M., \and Fabian, A.
389: C. 1991, \mnras, 249, 352
390:
391: \bibitem[Giommi \& Padovani(1994)]{giommi94}Giommi, P., \and Padovani,
392: P. 1994, \mnras, 268, 51
393:
394: \bibitem[Iwasawa \& Taniguchi(1993)]{iwasawa93}Iwasawa, K. \and
395: Taniguchi, Y. 1993, \apjl, 413, L15
396:
397: \bibitem[Jahoda et al.(2006)]{jahoda06}Jahoda, K., Markwardt, C. B.,
398: Radeva, Y., Rots, A. H., Stark, M. J., Swank, J. H., Strohmayer, T. E.,
399: \and Zhang, W., 2006, \apjs, 163, 401
400:
401: \bibitem[Jiang, Wang \& Wang(2006)]{jiang06}Jiang, P., Wang, J. X., \and
402: Wang, T. G. 2006, \apj, 644, 725
403:
404: \bibitem[Lamer, Uttley \& McHardy(2000)]{lamer00}Lamer, G., Uttley, P.,
405: \and McHardy, I. M. 2000, \mnras, 319, 949
406:
407: \bibitem[Lubi\'{n}ski \& Zdziarski(2001)]{lubinski01}Lubi\'{n}ski, P.,
408: \and Zdziarski, A. A. 2001, \mnras, 323, L37
409:
410: \bibitem[Magdziarz \& Zdziarski(1995)]{magdziarz95}Magdziarz, P. \&
411: Zdziarski, A. A. 1995, \mnras, 273, 837
412:
413: \bibitem[Markowitz, Edelson, \& Vaughan(2003)]{markowitz03}Markowitz,
414: A., Edelson, R., \and Vaughan, S. 2003, \apj, 598, 935
415:
416: \bibitem[Miniutti et al.(2006)]{miniutti06}Miniutti, et al. 2006, PASJ,
417: accepted (astro-ph/0609521)
418:
419: \bibitem[Page et al.(2004)]{page04}Page, K. L., O'Brien, P. T., Reeves,
420: J. N. \and Turner, M. J. L. 2004, \mnras, 347, 316
421:
422: \bibitem[Papadakis et al.(2002)]{papadakis02}Papadakis, I. E., Petrucci,
423: P. O., Maraschi, L., McHardy, I. M., Uttley, P., \and Haardt, F. 2002,
424: \apj, 573, 92
425:
426: \bibitem[Perola et al.(2002)]{perola02}Perola, G. C., Matt, G., Cappi,
427: M., Fiore, G., Guainazzi, M., Maraschi, L., Petrucci, P. O., \and Piro,
428: L. 2002, \aap, 389, 802
429:
430: \bibitem[Wang et al.(1999)]{wang99}Wang, J. X., Zhou, Y. Y., Xu, H. G.,
431: \and Wang, T. G. 1999, \apj, 516, L65
432:
433: \bibitem[Weaver, Gelbord \& Yaqoob(2001)]{weaver01}Weaver, K. A.,
434: Gelbord, J., \and Yaqoob, T. 2001, \apj, 550, 261
435:
436: \bibitem[Zdziarski, Lubi\'{n}ski \& Smith(1999)]{zdziarski99}Zdziarski,
437: A. A., Lubi\'{n}ski, P., \and Smith, D. A. 1999, \mnras, 303, L11
438:
439:
440:
441: \end{thebibliography}
442:
443: \clearpage
444:
445: \begin{deluxetable}{lrrrrr}
446: \tablecolumns{6}
447: \tablecaption{Sample of \RXTE-observed Seyfert 1 and 1.2 galaxies\tablenotemark{a}}
448: \tablewidth{0pt}
449:
450: \tablehead{
451: \colhead{Galaxy} &
452: \colhead{Seyfert} &
453: \colhead{Fitted} &
454: \colhead{Average} &
455: \multicolumn{2}{c}{$EW/L_x$ correlation\tablenotemark{e}}
456: \\
457: \cline{5-6}
458: \\
459: \colhead{} &
460: \colhead{Type\tablenotemark{b}} &
461: \colhead{Spectra\tablenotemark{c}} &
462: \colhead{$\sigma_{Fe K\alpha}$\tablenotemark{d}} &
463: \colhead{$\alpha$} &
464: \colhead{WV/Num.}
465: }
466:
467: \startdata
468: \medskip
469: All & & & & -0.14$^{+0.01}_{-0.01}$ & 700/350
470: \\
471: \
472: \medskip
473: Quasars & & & & +0.09$^{+0.20}_{-0.25}$ & 105/81
474: \\
475: \smallskip \hspace{12pt}
476: 3C 273 & 1 & 81 & 0.329 & +0.09$^{+0.20}_{-0.25}$ & 105/81
477: \\
478:
479: \medskip
480: Broadline Seyfert 1s & & & & -0.24$^{+0.14}_{-0.15}$ & 48.0/66
481: \\
482: \smallskip \hspace{12pt}
483: 3C 111 & 1 & 4 & 0.239 & +0.70$^{+2.60}_{-1.52}$ & 0.654/4
484: \\
485: \smallskip \hspace{12pt}
486: 3C 120\tablenotemark{f}& 1 & 40 & 0.261 & -0.70$^{+0.63}_{-0.61}$ & 20.9/39
487: \\
488: \smallskip \hspace{12pt}
489: 3C 382 & 1 & 5 & 0.328 & -0.80$^{+1.69}_{-1.70}$ & 2.54/5
490: \\
491: \smallskip \hspace{12pt}
492: 3C 390.3 & 1 & 17 & 0.203 & -0.51$^{+0.44}_{-0.41}$ & 2.70/17
493: \\
494:
495: \medskip
496: Seyfert 1s (Radio quiet)& & & & 0.01$^{+0.30}_{0.30}$ & 23.6/31
497: \\
498: \smallskip \hspace{12pt}
499: Ark 120 & 1 & 15 & 0.197 & -0.66$^{+0.58}_{-0.57}$ & 6.62/15
500: \\
501: \smallskip \hspace{12pt}
502: Fairall 9 & 1 & 16 & 0.155 & +0.41$^{+0.44}_{-0.44}$ & 11.1/16
503: \\
504:
505: \medskip
506: Seyfert 1.2s & & & & -0.08$^{+0.03}_{-0.03}$ & 192/169
507: \\
508: \smallskip \hspace{12pt}
509: IC 4329A & 1.2 & 41 & 0.214 & -0.55$^{+0.36}_{-0.37}$ & 27.5/41
510: \\
511: \smallskip \hspace{12pt}
512: MCG -6-30-15 & 1.2 & 75 & 0.292 & -0.65$^{+0.34}_{-0.33}$ & 89.2/75
513: \\
514: \smallskip \hspace{12pt}
515: Mkn 509 & 1.2 & 16 & 0.102 & -0.52$^{+0.91}_{-0.99}$ & 7.57/16
516: \\
517: \smallskip \hspace{12pt}
518: NGC 7469 & 1.2 & 37 & 0.145 & -0.58$^{+0.30}_{-0.31}$ & 17.7/37
519: \\
520:
521: \medskip
522: Narrow Line Seyfert 1 & & & & 8.80$^{+20.80}_{-6.08}$ & 0.196/3
523: \\
524: \smallskip \hspace{12pt}
525: TON S180 & 1.2 & 3 & 0.379 & 8.80$^{+20.80}_{-6.08}$ & 0.196/3
526:
527: \enddata
528:
529: \tablenotetext{a}{The following sources were eliminated after running the
530: data pipeline described in the text, due to having no spectra with at
531: least 125,000 net photons: Mkn 110, PG 0804+761, PG 1211+143, Mkn 79,
532: Mkn 335, and PG 0052+251.}
533: \tablenotetext{b}{Seyfert type based on the NASA Extragalactic Database}
534: \tablenotetext{c}{Total number of spectra extracted using our data pipeline (\S~\ref{subsection:pipeline}).}
535: \tablenotetext{d}{The average physical width of the Fe K$\alpha$ line
536: for all spectra from a source when fitted to the absorbed powerlaw model
537: with Compton reflection and Gaussian iron line
538: (\S~\ref{subsection:results}).}
539: \tablenotetext{e}{Results of fitting the X-ray luminosity over EW plot
540: to a powerlaw model; e.g. $EW \propto L_x^{\alpha}$, where $L_x$ is the
541: 2-10 keV X-ray luminosity in ergs s$^{-1}$ and $EW$ is the iron line
542: equivalent width in eV.}
543: \tablenotetext{f}{One 3C 120 spectrum shows a flare, where $L_x$
544: jumps by $\sim6\times$. The number quoted above excludes this point
545: from the sample. If we include the flare, we find $EW \propto
546: L_x^{0.07(+0.18/-0.25)}$.}
547:
548: \label{table:sources}
549:
550: \end{deluxetable}
551:
552: \clearpage
553:
554: \begin{figure}
555: \plottwo{f1a.eps}{f1b.eps}
556: \caption{Iron line equivalent width in eV ($EW$) versus powerlaw photon
557: index ($\Gamma$) for the Seyfert 1/1.2 sample (a) and for the Monte Carlo
558: simulations (b).}
559: \label{fig1}
560: \end{figure}
561:
562: \clearpage
563:
564: \begin{figure}
565: \plottwo{f2a.eps}{f2b.eps}
566: \caption{Reflection fraction ($R$) versus powerlaw photon index ($\Gamma$) for
567: the Seyfert 1/1.2 sample (a) and for the Monte Carlo simulations
568: (b). In both plots, the line shows the best-fit linear model for
569: the Monte Carlo simulations.}
570: \label{fig2}
571: \end{figure}
572:
573: \clearpage
574:
575:
576: \begin{figure}
577: \plotone{f3.eps}
578: \caption{The iron line equivalent width in eV versus the powerlaw
579: photon index. This plot is similar to the left panel in
580: Figure~\ref{fig1}, but with each galaxy plotted with a separate
581: symbol. The open circles are 3C 111, open squares are 3C120,
582: pluses (+) are 3C273, open triangles are 3C 382, open diamonds 3C
583: 390.3, open stars Akn 120, open crosses Fairall 9, filled circles
584: IC 4329A, filled squares MCG $-$6-30-15, filled triangles Mkn 509,
585: filled stars NGC 7469, and asterisks (*) TON S180.}
586: \label{fig3}
587: \end{figure}
588:
589: \end{document}
590:
591:
592:
593: