0707.0508/ms.tex
1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2004 January 9
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8: 
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12: 
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: \documentclass[12pt,preprint]{aastex}
19: 
20: %% manuscript produces a one-column, double-spaced document:
21: 
22: %% \documentclass[manuscript]{aastex}
23: 
24: %% preprint2 produces a double-column, single-spaced document:
25: 
26: %% \documentclass[preprint2]{aastex}
27: 
28: %% Sometimes a paper's abstract is too long to fit on the
29: %% title page in preprint2 mode. When that is the case,
30: %% use the longabstract style option.
31: 
32: %% \documentclass[preprint2,longabstract]{aastex}
33: 
34: %% If you want to create your own macros, you can do so
35: %% using \newcommand. Your macros should appear before
36: %% the \begin{document} command.
37: %%
38: %% If you are submitting to a journal that translates manuscripts
39: %% into SGML, you need to follow certain guidelines when preparing
40: %% your macros. See the AASTeX v5.x Author Guide
41: %% for information.
42: 
43: \newcommand{\kev}{\mbox{$\rm\,keV$}}
44: \newcommand{\cm}{\mbox{$\rm\,cm$}}
45: %\newcommand{\vdag}{(v)^\dagger}
46: %\newcommand{\myem%
47: %\slugcomment{Not to appear in Nonlearned J., 45.}
48: 
49: %% If you wish, you may supply running head information, although
50: %% this information may be modified by the editorial offices.
51: %% The left head contains a list of authors,
52: %% usually a maximum of three (otherwise use et al.).  The right
53: %% head is a modified title of up to roughly 44 characters.
54: %% Running heads will not print in the manuscript style.
55: 
56: %% You can insert a short comment on the title page using the command below.
57: 
58: \slugcomment{Submitted to ApJ on 12/Jun/2006}
59: 
60: %% If you wish, you may supply running head information, although
61: %% this information may be modified by the editorial offices.
62: %% The left head contains a list of authors,
63: %% usually a maximum of three (otherwise use et al.).  The right
64: %% head is a modified title of up to roughly 44 characters.
65: %% Running heads will not print in the manuscript style.
66: 
67: \shorttitle{Spectral Signatures of Photon Bubbles}
68: \shortauthors{Finke \& B\"ottcher}
69: 
70: %% This is the end of the preamble.  Indicate the beginning of the
71: %% paper itself with \begin{document}.
72: 
73: \begin{document}
74: 
75: %% LaTeX will automatically break titles if they run longer than
76: %% one line. However, you may use \\ to force a line break if
77: %% you desire.
78: 
79: \title{X-ray Spectral Signatures of the Photon Bubble Model for 
80: Ultraluminous X-ray Sources}
81: 
82: 
83: %% Use \author, \affil, and the \and command to format
84: %% author and affiliation information.
85: %% Note that \email has replaced the old \authoremail command
86: %% from AASTeX v4.0. You can use \email to mark an email address
87: %% anywhere in the paper, not just in the front matter.
88: %% As in the title, use \\ to force line breaks.
89: 
90: \author{Justin D. Finke and Markus B\"ottcher}
91: \affil{Astrophysical Institute, Department of Physics and Astronomy, \\
92: Ohio University, Athens, OH 45701}
93: \email{finke@helios.phy.ohiou.edu, mboett@helios.phy.ohiou.edu}
94: 
95: %% Notice that each of these authors has alternate affiliations, which
96: %% are identified by the \altaffilmark after each name.  Specify alternate
97: %% affiliation information with \altaffiltext, with one command per each
98: %% affiliation.
99: 
100: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
101: %% style, abstract will output a Received/Accepted line after the
102: %% title and affiliation information. No date will appear since the author
103: %% does not have this information. The dates will be filled in by the
104: %% editorial office after submission.
105: 
106: \begin{abstract}
107: The nature of ultraluminous X-ray sources in nearby galaxies is one of
108: the major open questions in modern X-ray astrophysics. One possible 
109: explanation for these objects is an inhomogeneous, radiation dominated 
110: accretion disk around a $\sim 10 M_{\odot}$ black hole --- the so-called
111: ``photon bubble'' model. While previous studies of this model have 
112: focused primarily on its radiation-hydrodynamics aspects, in this
113: paper, we provide an analysis of its X-ray spectral (continuum and possible
114: edge and line) characteristics. Compton reflection between high and 
115: low density regions in the disk may provide the key to distinguishing 
116: this model from others, such as accretion onto an intermediate mass black
117: hole. We couple a Monte Carlo/Fokker-Planck radiation transport 
118: code with the \-XSTAR code for reflection to simulate the 
119: photon spectra produced in a photon bubble model for ULXs.  
120: We find that reflection components tend to be very weak and 
121: in most cases not observable, and make predictions for the shape 
122: of the high-energy Comptonizing spectra.  In many cases the 
123: Comptonization dominates the spectra even down to $\sim$ a few 
124: keV.  In one simulation, a $\sim 9 \kev$ feature was found, 
125: which may be considered a signature of photon bubbles in ULXs; 
126: furthermore, we make predictions of high energy power-laws 
127: which may be observed by future instruments.  
128: 	
129: 
130: \end{abstract}
131: 
132: %% Keywords should appear after the \end{abstract} command. The uncommented
133: %% example has been keyed in ApJ style. See the instructions to authors
134: %% for the journal to which you are submitting your paper to determine
135: %% what keyword punctuation is appropriate.
136: 
137: %% Authors who wish to have the most important objects in their paper
138: %% linked in the electronic edition to a data center may do so in the
139: %% subject header.  Objects should be in the appropriate "individual"
140: %% headers (e.g. quasars: individual, stars: individual, etc.) with the
141: %% additional provision that the total number of headers, including each
142: %% individual object, not exceed six.  The \objectname{} macro, and its
143: %% alias \object{}, is used to mark each object.  The macro takes the object
144: %% name as its primary argument.  This name will appear in the paper
145: %% and serve as the link's anchor in the electronic edition if the name
146: %% is recognized by the data centers.  The macro also takes an optional
147: %% argument in parentheses in cases where the data center identification
148: %% differs from what is to be printed in the paper.
149: 
150: \keywords{accretion, accretion disks --- methods:  numerical --- 
151: radiative transfer --- X-rays:  binaries}
152: 
153: %individual(\objectname{NGC 6397},
154: %\object{NGC 6624}, \objectname[M 15]{NGC 7078},
155: %\object[Cl 1938-341]{Terzan 8})}
156: 
157: %% From the front matter, we move on to the body of the paper.
158: %% In the first two sections, notice the use of the natbib \citep
159: %% and \citet commands to identify citations.  The citations are
160: %% tied to the reference list via symbolic KEYs. The KEY corresponds
161: %% to the KEY in the \bibitem in the reference list below. We have
162: %% chosen the first three characters of the first author's name plus
163: %% the last two numeral of the year of publication as our KEY for
164: %% each reference.
165: 
166: 
167: \section{Introduction}
168: \label{intro}
169: 
170: Approximately 150 off-nuclear Ultraluminous X-ray Sources (ULXs) in
171: nearby galaxies have been discovered with luminosities greater than 
172: $10^{39}$ erg s$^{-1}$, 
173: exceeding the Eddington Luminosity for
174: a $\sim 10 M_{\odot}$ black hole \citep[e.g.,][]{fabbiano88,cm99,cp02,lm05}.  
175: While some can be identified as supernova remnants, background Active Galactic
176: Nuclei, or faint foreground stars \citep[e.g.,][]{g06}, most seem to be the 
177: result of accretion from a high-mass star onto a compact object. Short-term
178: variability of some of these ULXs indicates that they are not simply 
179: unresolved superpositions of several lower-luminosity sources 
180: \citep[e.g.,][]{matsumoto01,fabbiano03}. Thus, the high 
181: luminosity implies accretion onto black holes with masses 
182: $50 M_{\odot} < M < 10^4 M_{\odot}$,
183: \citep[intermediate mass black holes; IMBHs;][]{cm99,metal00,hui05,metal06} 
184: or super--Eddington accretion.  Super--Eddington accretion can be achieved
185: in two different ways:  an inhomogeneous accretion-disk structure in which 
186: the photon flux is spatially separated from the bulk of the matter influx, 
187: or strongly anisotropic radiation. 
188: 
189: Anisotropic emission \citep{king01} may originate, e.g., from jet sources
190: associated with accretion onto solar-mass black holes (microquasars)
191: in which the jet is oriented at a small angle with respect to our line
192: of sight --- the so-called ``microblazar'' model \citep{gak02,kfm02}.   
193: The latter model is now considered unlikely based on recent observations 
194: of X-ray ionization of optical nebulae associated with some ULXs 
195: \citep{pm03, g06}.
196: 
197: A promising mechanism for sustaining 
198: super-Eddington accretion in a radiation-dom\-in\-at\-ed, magnetized accretion 
199: flow, is provided by the so-called photon bubble instability 
200: \citep{a92,g98,begelman01,begelman02,begelman06}. 
201: In this model, radiation can escape 
202: along low-density regions (LDRs) aligned with predominantly vertical
203: magnetic field lines.  The accretion flow is concentrated in thin,
204: optically thick high-density regions (HDRs), where the magnetic pressure
205: of field lines oriented predominantly within the disk dominates over 
206: the gas pressure, thus confining the gas and preventing the radiation
207: pressure from disrupting the accretion flow. In such a configuration,
208: the total disk luminosity can exceed the Eddington luminosity by a
209: factor approximately equal to the ratio of the magnetic field pressure
210: to the gas pressure \citep{begelman01,begelman02} in the HDRs. 
211: 
212: Given the variety of different promising candidate models for the nature
213: of ULXs, the obvious question is: Can one distinguish between these models 
214: with ULXs' X-ray spectra? If so, what are the characteristic spectral features
215: of the individual models? From an accreting 
216: IMBH it is generally believed that a multi-color 
217: disk blackbody (MCDBB) spectrum with an inner temperature of $kT_{in} \sim$ 
218: 0.1 --- 0.3~keV (that is, the disk temperature at the innermost stable 
219: circular orbit) and possibly a high energy component 
220: from Compton up-scattering 
221: in a tenuous, hot corona would be a realistic phenomenological description
222: of the spectrum. However, a recent, more detailed analysis of non-LTE
223: accretion flows around IMBHs \citep{hui05} has indicated that the effects
224: of black-hole rotation and Compton scattering within the disk may very well
225: lead to much higher apparent disk temperatures, up to $kT \sim 1$~keV,
226: in addition to deviations from conventional MCDBB spectra at both soft
227: and hard X-ray energies due to metal opacity effects. 
228: 
229: The continuum spectrum from the photon bubble model might be dominated 
230: by the MCDBB spectrum emanating from the
231: HDRs.
232: However, this spectrum might be modified during the 
233: radiation transport in the photon bubble cavities. Furthermore, the almost 
234: free-streaming radiation of the photon bubbles could be repeatedly 
235: Compton-reflected off the surfaces of the HDRs, 
236: potentially leading to strong fluorescence lines and/or radiative 
237: recombination edges, in addition to Compton reflection features from 
238: hard X-ray emission impinging upon the disk from radiation sources
239: external to the disk \citep{btb04,bty05}. Such features may be 
240: observable, assuming it is not overwhelmed by other radiation sources, 
241: in particular the blackbody from the HDRs. 
242: 
243: Distinguishing scenarios by spectral modeling is currently very difficult 
244: for all but the highest quality ULX data sets, and spectral fitting to 
245: these ULX spectra gives contradictory results. For example, \citet{fk05}
246: recently performed a detailed spectral and timing analysis of archival 
247: {\it XMM-Newton} data on 28 ULXs that had sufficient photon 
248: statistics to allow for meaningful fitting with models more 
249: complicated than a simple power-law.  They found that their continuum 
250: spectra fell into three general categories: (1) Optically thin 
251: bremsstrahlung-dominated, quasi-thermal spectra with temperatures 
252: of $kT \sim 0.6$ -- 0.8~keV, characteristic for X-ray emission from 
253: young supernova remnants; (2) MCDBB spectra of temperatures $kT \sim 
254: 0.1$ -- 0.4~keV, plus occasionally a hard X-ray power-law, as possible 
255: in the case of accretion onto IMBHs; (3) MCDBB spectra at temperatures 
256: $kT \sim 1$~keV, plus a power-law component dominating at lower energies. 
257: Based on these results, 
258: \cite{fk05} suggest that ULXs may, in fact, not be a homogeneous 
259: class of objects. This is 
260: in accord with recent results of \cite{metal06}
261: that realistic stellar evolution and population synthesis calculations
262: suggest that the expected rate of captures of massive stars by IMBHs
263: may not be sufficient to produce the total observed number of ULXs. 
264: \citet{wetal06} classified ULXs into low/hard and high/soft 
265: states based on their X-ray luminosities and spectra, assuming the
266: ULXs were IMBHs accreting in states similar to galactic X-ray binaries.
267: They found that the high/soft ULXs were grouped around
268: one of two blackbody temperatures:  one grouped around $~1$ keV and one
269: grouped around $~0.1$ keV, further indication that ULXs are
270: not a homogeneous group. The long term monitoring of some ULXs such
271: as NGC 5204 X-1 shows that its X-ray spectrum hardens as its flux
272: increases, opposite to what has been observed in Galactic X-ray
273: binaries \citep{robetal06}.
274: 
275: \citet{srw06} found that 6 out of 13 of the most-observed ULXs were 
276: fit approximately equally as well with a cool MCDBB and hard power-law 
277: as with a soft power-law and warm MCDBB.  Even more surprisingly, 
278: they found that most spectra (10 out of 13) were fit best by a cool, 
279: $kT \sim 0.2$~keV blackbody and a warmer $kT_{in} \sim 2$~keV MCDBB, 
280: and 11 out of 13 were fit by a cool MCDBB and an optically thick 
281: ($\tau \sim 10$) corona. These spectra could also be explained by 
282: a hot, inner (optically thick) plasma sphere and a cool outer disk 
283: \citep{am06}, in which 
284: case the cool blackbody temperature could not be directly related 
285: to the black hole mass by virtue of the relation between the inner 
286: disk temperature and the black hole mass. 
287: 
288: In summary, current diagnostics for the distinction between different
289: models for ULXs indicate that ULXs may in fact not be a homogeneous
290: set of physical objects, different models may apply to different
291: sources, and in many cases, observational results are inconclusive. 
292: Consequently, more detailed model predictions of various ULX models
293: might be helpful in identifying additional diagnostics which may be
294: used to confirm or rule out such models. In this paper, we investigate
295: detailed spectral features resulting from the photon bubble model,
296: both in the continuum and possible emission line and radiative
297: recombination edge features resulting from the radiative feedback
298: between the HDRs and LDRs. An Fe K line has 
299: been observed in M82 X-1 \citep{sm03,am06}; absorption edges have 
300: been observed in the 0.1 -- 1~keV range in M101 ULX-1 \citep{ketal04}; 
301: and possible edges at $\sim 0.7$ keV and $\sim 8$ keV, and an
302: emission feature at $\sim 6$~keV has been seen in Holmberg IX
303: X-1 \citep{dgr06}.  In the ULX M51 X-26 emission features
304: have been seen at 1.8~keV, 3.24~keV, 4.03~keV, and 6.65~keV with 
305: {\it Chandra} \citep{tw04} and at 6.4~keV with {\it XMM--Newton}
306: \citep{detal05}. This indicates that the detection of emission
307: lines and/or radiative recombination edges is currently feasible,
308: at least for bright ULXs, and predictions of expected line features
309: from various models will thus be useful as an additional model
310: diagnostic.  
311: 
312: Studies of X-ray reflection in inhomogeneous accretion disks
313: have been done by
314: \citet{rfb02} and \citet{fetal02}, who found that multiple 
315: Compton reflections of emission from a corona above
316: an irregular or corrugated accretion disk in Seyfert galaxies
317: lead to much stronger reflection features than single reflections.  
318: \citet{meretal06} considered
319: Compton reflection in a radiation--dominated inhomogeneous
320: accretion disk.  Their disk was a low density Comptonizing plasma with
321: high density clumps off of which radiation was reflected.  Their 
322: simulations produced lines which may be visible in narrow line Seyfert 
323: 1 galaxies, although detection may be difficult due to 
324: interstellar  absorption. 
325: So far, investigations of photon bubbles in ULXs have focused
326: on radiation--hydrodynamics \citep[see, e.g., ][]{rb03, tetal05} 
327: leaving the expected spectral features essentially unexplored.
328: \citet{btb04} and \citet{bty05} studyed X-ray reflection off accretion
329: disks in the photon bubble model for Seyfert galaxies, but using
330: an unspecified external irradiating hard X-ray source, such as a 
331: tenuous, hot corona above the accretion disk surface, while we are 
332: focusing on the self-consistent local radiation feedback between the 
333: HDRs and LDRs within the inhomogeneous accretion flow. 
334: Furthermore, those authors did not explore parameters
335: appropriate for ULXs. 
336: 
337: In this paper we make predictions of the X-ray
338: spectra from the photon bubble model of ULXs with a 
339: Monte Carlo/Fokker-Planck code coupled with the XSTAR program for 
340: X-ray reprocessing in the HDRs (reflection).
341: We will be primarily interested in predictions
342: of spectral features in addition to a warm MCDBB and soft power-law
343: component, although this is not a study of any individual object.  
344: In addition we make predictions of a power-law component 
345: that extends above 10 keV which may not be detectable by 
346: {\em Chandra} or {\em XMM-Newton} but may be by {\em Suzaku} or 
347: future telescopes.  
348: 
349: \section{Model Setup}
350: \label{setup}
351: 
352: The simulation parameters were based on solutions found in 
353: \citet{begelman01} and \citet{begelman02}.  
354: We summarize these solutions in \S\ \ref{disk},
355: and describe the simulation technique in \S\ \ref{simulation}.  
356: 
357: \subsection{Disk Distribution}
358: \label{disk}
359: 
360: The physical picture emerging from numerical simulations of the
361: photon bubble instability \citep{rb03,tetal05} points towards
362: a propagating pattern of slab-like shock trains (the 
363: HDRs) slanted with respect to the plane of the accretion disk.
364: A solution to the disk structure has been derived by
365: \citet{begelman02}, who found the relevant disk parameters to depend on:
366: the accretion rate, $\dot{m} = \dot{M}\kappa c/4\pi GM_{BH}$, 
367: $\beta$, the ratio of the HDR's gas pressure 
368: to the magnetic pressure, and $\alpha$, the Shakura-Sunyaev viscosity
369: parameter.  Here 
370: $\kappa$ is the opacity (assumed to be dominated by 
371: Thomson scattering; $\kappa=$
372: 0.4 cm$^2$ g$^{-1}$), $c$ is the
373: speed of light, $G$ is the gravitational constant, and $M_{BH}$ is 
374: the mass of the black hole; We used his solution to set up our simulations.
375: Throughout this work we assumed $\beta = 0.1$ and 
376: $m = M_{BH} / M_{\odot} = 10$ (resulting in an Eddington luminosity of 
377: $1.3 \cdot 10^{39}$ erg s$^{-1}$).
378: 
379: The Eddington enhancement factor, the ratio of the disk's flux
380: to the Eddington flux, is:
381: \begin{equation}
382: \label{ratio}
383: l \equiv \frac{F}{F_{Edd}} = \frac{n_{HDR}}{n_{avg}} 
384: = \frac{3}{2} \frac{\dot{m}D}{\delta r}
385: \end{equation}
386: where $n_{avg} = \sqrt{n_{HDR}n_{LDR}}$ is the geometric 
387: average of the LDR and HDR densities; $r$ is the distance from the
388: black hole in terms of gravitational radii ($R_g = GM/c^2$); 
389: $\delta$ is the ratio of the height to the radius ($\delta = h/R$); 
390: and $D =  1 - (6/r)^{1/2}$.  \citet{begelman02} derives the following
391: equation for $\delta$:
392: $$
393: \delta \sim  \max \Biggl[  0.3  \left( \frac{\beta}{0.1} \right) ^{-4/13} 
394: \left( \frac{\alpha}{0.01} \right) ^{-5/13} 
395: (\dot{m}D)^{5/13} m^{-1/13} r^{-5/26}, 
396: $$
397: \begin{equation}
398: \label{delta}
399: %\begin{eqnarray}
400:  0.2   \left( \frac{\beta}{0.1} \right)^{4/21} 
401: \left( \frac{\alpha}{0.01} \right)^{-5/21}
402: (\dot{m}D)^{5/21} m^{-1/21} r^{1/14} \Biggr]
403: %\end{eqnarray}
404: \end{equation}
405: Note that in Eqn. \ref{delta} 
406: we corrected a typographical error in Eqn. 13 of
407: \citet{begelman02}.  
408: Based on \citet{begelman02}'s expression for the average density, and
409: equation \ref{ratio}, the following equations were derived for
410: the densities in the HDR and LDR:
411: \begin{equation}
412: \label{HDR}
413: n_{HDR} = 6.0\cdot10^{18}(\dot{m}D)^2(m\alpha)^{-1}\delta^{-4}r^{-5/2}\ 
414:           \cm^{-3}
415: \end{equation}
416: \begin{equation}
417: \label{LDR}
418: n_{LDR} = 3.6\cdot10^{18}(m\alpha)^{-1}\delta^{-2}r^{-1/2}\ 
419:           \cm^{-3}.
420: \end{equation}
421: Although these densities refer to the disk midplane, the \citet{ss73} 
422: solution for a radiation-dominated accretion disk shows that the 
423: density will be essentially constant in the vertical direction.  
424: Thus, we consider these densities to be the density at the 
425: photosphere.
426: The temperature for the HDRs was found by assuming that all the 
427: gravitational potential energy
428: is released in the HDRs, and thus, the flux emitted at the 
429: photoplane is set equal to the blackbody
430: flux.  This gives the \citet{ss73} expression:
431: \begin{equation}
432: \label{tempHDR}
433: kT_{HDR} = 5.09\ m^{-1/4} (\dot{m}D)^{1/4} r^{-3/4}\ \kev.
434: \end{equation}
435: 
436: The ``wavelength'' of the plane-parallel shocks---i.e., this 
437: distance between HDRs---is given by 
438: \begin{equation}
439: \lambda = \beta\alpha\delta R.  
440: \end{equation}
441: 
442: 
443: For the proton temperature in the LDRs, the virial temperature 
444: was used:
445: \begin{equation}
446: \label{kTp}
447: kT_{lp} = \frac{GM_{BH}m_p}{R}
448: \end{equation}
449: 
450: where $m_p$ is the proton's mass.  
451: 
452: The electron temperature in the LDR is expected to be significantly
453: lower than the proton temperature due to Compton cooling by radiation
454: from the HDRs.  The LDR's electron temperature
455: was determined by implicitly solving a
456: Fokker-Planck equation with the MC/FP code; 
457: see \S\ \ref{simulation} below.  The magnetic field in the 
458: LDR was determined by assuming it to be in equipartition with the
459: electrons.
460: 
461: The Eddington ratio is dependent on radius (equation \ref{ratio}), 
462: so for a given disk, outside of a certain radius ($r_{inhom}$) 
463: $l<1$, and the 
464: disk becomes a homogeneous, Shakura-Sunyaev disk.  Outside of this
465: radius, the disk was represented by MCDBB with
466: $kT \propto r^{-3/4}$.
467:  
468: \subsection{Simulation Description}
469: \label{simulation}
470: 
471: Equations \ref{delta} through \ref{kTp} 
472: were used to determine
473: the simulation parameters at various radii.  To find the total spectrum
474: of the disk, simulations were run
475: at evenly-spaced radii and 
476: the results were averaged, weighted by disk area.  
477: At each radius, we simulated one individual LDR 
478: sandwiched by two HDRs in a plane-parallel geometry 
479: (see Fig. \ref{geometry}).  Within the plane-parallel geometry the 
480: LDR is divided into 40 zones, 4 radial and 10 vertical.
481: It seems reasonable to assume, that, since 
482: $\lambda$ is much smaller than the radiation 
483: pressure scale height \citep{begelman02}, that the fraction of photons which 
484: escape (given by the ratio of the escape area to the region's 
485: total surface area), will be $\sim 1$\%.  Larger escape fractions 
486: could lead to smaller ionization parameters and greater reflection components,
487: however, this was not explored in this work.  
488: Multiple Compton reflections are certainly possible with such
489: a small escape fraction, however for nearly all of our cases, 
490: the reflected component was completely drowned out by the 
491: blackbody; see below.  As a result, further iterations did not 
492: produce significantly different spectra.  
493: 
494: The radiation transfer and self-consistent
495: balance between electron heating and cooling within the LDR are 
496: simulated with the two-dimensional
497: Monte Carlo/Fokker-Planck
498: (MC/FP) code described in \citet{bjl03} and \citet{bl01}.  
499: This code 
500: uses the Monte Carlo method of \citet{pss83} for Compton scattering
501: and the implicit Fokker-Planck method of \citet{nm98} for the
502: evolution
503: of the electron distribution in a two-temperature plasma with a 
504: given proton temperature.  In each zone, 
505: the Fokker-Planck equation is implicitly and independently solved.  
506: The Fokker-Planck technique takes into
507: account heating/cooling by Coulomb/M\o ller interactions, Compton scattering,
508: synchrotron/cyclotron processes.  
509: 
510: Initially, the emission emanating from the HDRs was represented by 
511: blackbody spectra inserted at the upper and lower boundaries.  
512: Photons were also produced by synchrotron/cyclotron processes
513: in the LDRs.  Photons were subjected to Compton scattering and 
514: reflection off the HDRs.
515: Escaping photons at the outer boundaries were added to an event 
516: file for later spectral extraction; this involves
517: placing each escaping photon in a particular energy bin.  
518: 
519: The proton temperature in the LDR was calculated from equation
520: \ref{kTp}, and the electron temperature
521: was calculated numerically within 
522: the MC/FP simulation \citep[see][]{nm98,bjl03,bl01,fb05}, which was run 
523: until the electron temperature reached a stable equilibrium.
524: The MC/FP
525: simulation was run to extract the photon flux and spectrum 
526: incident on the boundaries between the HDRs and the LDR. To 
527: calculate the expected spectral features from fluorescence
528: line emission, radiative recombination, and Compton reflection, 
529: the impinging spectrum was used as an input into the latest 
530: version of XSTAR \citep{kb01}. Solar abundances of the most 
531: profuse astrophysical elements (H, He, C, Ca, N, O, Ne, Mg, 
532: Si, S, Ar, Ca and Fe) were assumed in the HDR, based on \citet{gns96}.
533: XSTAR was run in constant pressure mode; the pressure was calculated 
534: from the specified density and the ideal gas law. 
535: The main parameter in determining the reflection spectrum's shape 
536: and intensity is the ionization parameter:
537: \begin{equation}
538: \label{ionparam}
539: \xi = \frac{4 \pi F}{ n_{HDR} }.
540: \end{equation}
541: Note that XSTAR uses the flux calculated between 1 and 1000 Ry 
542: (13.6 eV to 13.6 keV).
543: In order to 
544: circumvent XSTAR's limitation to densities $n_e \lesssim 
545: 1 \cdot 10^{17}$cm$^{-3}$, the impinging flux was re-scaled
546: to keep the ionization $\xi$ at the 
547: value corresponding to the physical situation. 
548: The inverse flux
549: scaling was applied to the XSTAR output spectrum.  Keeping 
550: $\xi$ constant insures that the scaling of a flux dominated by
551: recombination features, $F \propto n_e$ is properly recovered.  
552: Testing of XSTAR with various densities but the same $\xi$ 
553: seems shows 
554: that this is appropriate for the energy range of interest.   
555: The resulting 
556: reflection spectrum was then added to the intrinsic blackbody 
557: from the HDR as boundary sources in a second run of the MC/FP 
558: code for the final evaluation of the emanating X-ray spectrum.
559: For an example of a simulation, with spectra at different
560: radii, see Fig. \ref{specradii}.
561: 
562: 
563: \section{Results}
564: \label{results}
565: 
566: The simulation parameters and fit results 
567: can be seen in Table \ref{parameters}.
568: The important disk parameters and are plotted as a function of radius
569: in Figs. \ref{paramradius1} to \ref{paramradius3}.  
570: 
571: The transition between the expressions for $\delta$ (Eqn. \ref{delta})
572: can be readily seen in panel (a) of Fig. \ref{paramradius1},
573: with the first expression being used at lower radii.  At higher
574: $\alpha$, the transition occurs at lower radii, or not at all.
575: Note that several simulations violate the thin-disk condition 
576: ($\delta < 1$), especially the high $\dot{m}$ and/or low
577: $\alpha$ ones.  These solutions should thus be viewed with caution.
578: 
579: The ionization parameter decreases with radius, as 
580: is observed in our simulations, seen in panel (b) of Figs. 
581: \ref{paramradius1} to \ref{paramradius3}.  
582: At no point does $\xi$ drop below $\sim 2000$ erg cm s$^{-1}$, 
583: so the reflection component will be small ($\la 12\%\ $
584:  of the total) for all cases.  
585: 
586: The optical depth, $\tau$, is proportional to $\lambda n_{LDR}$, and is 
587: essentially always increasing with $r$.  Thus the spectra have 
588: more luminosity from Comptonization at larger radii.  It 
589: is also larger for smaller accretion rates due to its $\delta$
590: dependence 
591: ($\tau \propto \lambda n_{LDR} \propto \delta n_{LDR} \propto \delta^{-1}$)
592: by virtue of $\delta$  increasing with increasing $\dot{m}$.
593: Thus we expect a greater Comptonization component for 
594: smaller $\dot{m}$.
595: We can see that for our simulations, this is in fact the case 
596: (Fig. \ref{spectot}).  We can also see in 
597: Table \ref{parameters} that the fractional power emitted through
598: Comptonization decreases with increasing $\dot{m}$, as expected.  
599: This also corresponds to the decrease in $\Gamma$, the 
600: photon index ($EL_E \propto E^{-\Gamma+2}$) with increasing 
601: $\dot{m}$.  One can also see in Figs. 
602: \ref{paramradius1}---\ref{paramradius3} that as $\alpha$ increases, 
603: $\tau$ increases.  Again, this is because 
604: $\tau \propto \delta^{-1}$, and as $\alpha$ increases, $\delta$ 
605: decreases.  
606: 
607: The ratio $l$ peaks at $r\sim 12$,
608: as can be seen in panel (d) of
609: Figs. \ref{paramradius1}-- \ref{paramradius3}.  This is 
610: close to where the peak flux is expected in a typical 
611: Shakura-Sunyaev accretion disk.  Also note that $l$ 
612: increases with increasing $\alpha$, as one would expect from 
613: eqn. \ref{ratio} ($l\propto \delta^{-1}$).
614: 
615: For each simulation, the total reflection component from XSTAR 
616: was summed up 
617: for all radii.  The soft X-ray spectra were fit with a MCDBB.
618: The MCDBB and reflection components were subtracted from the total
619: spectra revealing the Comptonization components.  This component was 
620: fit with a power-law and exponential cutoff 
621: ($EL_E \propto E^{-\Gamma+2}\ e^{-E/E_{cutoff}}$) above 25 keV.  
622: The results of this 
623: decomposition and fits are summarized in Table \ref{parameters};
624: the individual fit components for the $\alpha=0.01$ simulations
625: can be seen in Fig. \ref{reflect}.  
626: Note that these fit components are the sum of the
627: simulations at all radii.  
628: The only simulation that has a significant reflection 
629: feature visible in its spectrum is Sim. 2, with a blended 
630: Fe XXV/XXVI feature at $\sim 9\ \kev$ with an equivalent width of 
631: $\sim 640$ eV, and an edge at $\sim 0.87\ \kev$ (O K).  This 
632: simulation also has the highest fraction of its emission 
633: from the reflection component.  One can see in Fig. \ref{specradii} 
634: that the reflection component is strongest at $r=120$---280.  
635: 
636: In Fig. \ref{specradii}, one can see
637: that the Comptonization component gets stronger at 
638: larger radii, as one would expect with the increasing 
639: optical depth.  
640: The overall spectra are generally dominated by the 
641: high-energy Comptonization component, especially for higher
642: $\dot{m}$.  
643: The Comptonization component in many cases extends 
644: far down in photon energy into the $\sim$ few keV range 
645: and often dominates the total bolometric luminosity, 
646: even when the total spectrum appears to be dominated 
647: by a thermal soft X-ray component.
648: 
649: This component contributes to the spectra even 
650: at low energies, down to $\sim 1 \kev$, as seen in 
651: Fig. \ref{reflect}.  In Fig. \ref{specradii}, one can see
652: that the Comptonization component gets stronger at 
653: larger radii, as one would expect with the increasing 
654: optical depth.  This explains the significant Comptonization 
655: component to the lower energy part of the spectrum.  
656: Eventually, though, one gets to the radius where photon bubbles 
657: can no longer be sustained ($r_{inhom}$), and there is 
658: no longer a Comptonization component, as seen in $r>520$ in 
659: Fig. \ref{specradii}.  Results for other simulations are similar 
660: to Fig. \ref{specradii}.  As $\alpha$ increases, the 
661: $r_{inhom}$ increases, as can be seen in Figs. 
662: \ref{paramradius1}---\ref{paramradius3} and 
663: Table \ref{parameters}.  
664: 
665: Spectral pivoting can be seen in the all of the 
666: simulations in Fig. \ref{spectot}.  
667: This is due to the approximately
668: same amount of luminosity in the Compton components of the different 
669: simulations.  The energy dissipated in the LDRs is related to the 
670: proton temperature, which in our simulations does not depend on 
671: accretion rate.  This is not an unrealistic assumption, as one 
672: would not expect the protons to radiate significantly.  
673: However, when the LDRs become more optically thick 
674: the energy dissipation becomes more efficient due to multiple 
675: Compton scatterings.  
676: 
677: \section{Discussion}
678: \label{discussion}
679: 
680: Our simulations explore 
681: 0.5--10 keV luminosities up to $\sim 8 \cdot 10^{39}$ erg s$^{-1}$
682: which does not include some of the brightest ULXs.  For our 
683: setup, luminosities much greater than this will violate 
684: the thin disk condition, except possibly for the highest value 
685: of $\alpha$.  It does not seem likely that photon bubbles could 
686: support much greater luminosities.  
687: 
688: Our results are significantly different from previous, similar 
689: works.  \citet{meretal06} perform similar calculations for a 
690: low density accretion region with imbedded optically thick 
691: clumps and considered reflection between these clumps.  
692: They used ionization parameters comparable to ours 
693: ($\sim 3000$ erg cm s$^{-1}$, among others) and 
694: achieved much stronger reflection components with no 
695: $\sim 9 \kev$ feature.  
696: This is a consequence of the substantially lower temperature 
697: of the soft blackbody (ultraviolet) photons used as an 
698: input into the Comptonization scheme in their model:  their 
699: Comptonization component extends down to $\la 0.1 \kev$ 
700: energies (in contrast to $\sim$ a few keV in our case).  
701: For comparable $\xi$, calculated from the 1---1000 Ry flux 
702: the number of photons above the Fe K edge is thus much
703: smaller than in our case.  Therefore, heavy metals 
704: are almost fully ionized in most of our simulations, leading 
705: to weak or absent line features and a relatively weak Compton
706: reflection component.
707: \citet{btb04} 
708: explored reflections from a hot corona off of an inhomogeneous 
709: disk and found the spectrum can differ significantly from 
710: reflection off a homogeneous disk.  They also had stronger 
711: reflection features than our results, due to 
712: lower ionization parameters, and no $\sim 9\ \kev$ 
713: feature.  
714: 
715: \citet{fk05} and \citet{srw06} fit ULXs' X-ray spectra 
716: with various models; when fit with a MCDBB or MCDBB and 
717: low-energy blackbody, their fits had inner disk temperatures 
718: similar to ours.  However, we find that our model is unable to 
719: explain a soft-excess.  We note that no $\sim 9 \kev$ 
720: feature has been detected in the spectra of ULXs, although 
721: features at other energies have been detected 
722: \citep{sm03,am06,ketal04,dgr06}, although these 
723: features could originate from a wind rather than 
724: reflection features in the disk.  ULX observations 
725: above $\sim 10 \kev$ by, {\em Suzaku} or the next generation
726: of hard X-ray imaging instruments might be able to 
727: detect the hard power-laws we predict, although
728: we realize such observations would be difficult.  
729: {\em Suzaku} observations of two ULXs in NGC 1313 
730: did not detect any component above 10 keV above the background 
731: \citep{mizuno07}, although their MCDBB + power-law fits do seem 
732: to agree with our spectra, and the variability of these sources is 
733: much less than observed in Galactic black hole candidates.  
734: 
735: Unfortunately, our simulations do not reproduce 
736: a soft excess as has been observed in many ULXs.  It is 
737: possible that the soft excess originates from 
738: Compton downscattering of radiation by wind \citep[e.g.,][]{begelman01}, 
739: and so a 
740: lack of its production in our simulations does not 
741: disprove the photon bubble model for ULXs.  
742: 
743: If the excess is explained, our model could explain 
744: ULXs well fit with a $\sim 1$ keV blackbody.  
745: Higher energy observations ($\ga10$ keV) could 
746: determine this; if they found a stable hard X-ray power-law 
747: $\Gamma \sim$ 2---2.4 for a long period of time, this would be 
748: evidence for the photon bubble model in ULXs, due to the 
749: fact that a lower mass compact object would have less 
750: variability \citep{kalogera04}.  
751: A $\sim 9 \kev$ feature could also be considered a signature 
752: of photon bubbles in ULXs; it has not been found in any other 
753: simulation of accreting black holes.  
754: 
755: We note that our spectra are similar to the very high state 
756: of X-ray binaries such as GX 339-4 \citep{belloni06} and 
757: GRO J1655-40 \citep{saito06}, which have similar 
758: photon indices as found in our simulations.  
759: For smaller accretion rates, the photon bubble
760: model may be a viable model to explain the very high state 
761: of X-ray binaries.
762: 
763: Since this is an exploratory study, we have made a number of 
764: simplifying assumptions.  
765: We have assumed a plane-parallel geometry;
766: zones could be corrugated or vary in random ways, 
767: which could lead to multiple reflections, increasing the reflection
768: component \citep{fetal02}.  
769: We have assumed constant densities across
770: LDRs.  This may alter the Comptonizing region of the spectra, 
771: but probably would not lead to significant differences.
772: We have completely neglected several items.
773: Taking into account 
774: General Relativistic effects could gravitationally broaden 
775: the spectrum.  However, most of the emission originates at 
776: $r\sim12$, too far out to be greatly affected by GR.  
777: We have assumed $n_{HDR}$ and $n_{LDR}$ to be constant in the 
778: vertical direction, neglecting color corrections; 
779: this may be validated by the more detailed 
780: vertical solution of \citet{begelman06}, who found 
781: that the corrections should be minor.    
782: Time variability is beyond the scope of this study 
783: and is poorly understood, and thus would involve 
784: poorly constrained parameters.
785: It is also possible 
786: that the spectra could be modified by further Comptonization in 
787: a corona above the disk, which we have also neglected.  This 
788: would make the spectra even harder than they already are, 
789: possibly too hard to match observations.  However, recent 
790: magnetohydrodynamic simulations have shown difficulties in 
791: creating coronae \citep{hirose06} and the photon bubble 
792: model may be an alternative to the standard disk-corona 
793: geometry.  Future simulations could take advantage of 
794: more detailed analytic solutions \citep[e.g.,][]{begelman06} 
795: and hydrodynamic simulations \citep[e.g.,][]{tetal05} 
796: and could include reprocessing in a corona and/or 
797: a disk wind.
798: 
799: \acknowledgments{
800: This work was partially supported by NASA through {\it XMM-Newton}
801: GO grant no. NNG04GI50G and INTEGRAL theory grant NNG05GK59G, 
802: as well as by an allocation of computing time from the 
803: Ohio Supercomputer Center (OSC) via grant PHS0256-1.  
804: Simulations were run on the OSC Pentium 4 Cluster in Columbus, Ohio.
805: We thank the referee for many helpful suggestions which have 
806: improved this paper.  
807: 
808: %% To help institutions obtain information on the effectiveness of their
809: %% telescopes, the AAS Journals has created a group of keywords for telescope
810: %% facilities. A common set of keywords will make these types of searches
811: %% significantly easier and more accurate. In addition, they will also be
812: %% useful in linking papers together which utilize the same telescopes
813: %% within the framework of the National Virtual Observatory.
814: %% See the AASTeX Web site at http://www.journals.uchicago.edu/AAS/AASTeX
815: %% for information on obtaining the facility keywords.
816: 
817: %% The reference list follows the main body and any appendices.
818: %% Use LaTeX's thebibliography environment to mark up your reference list.
819: %% Note \begin{thebibliography} is followed by an empty set of
820: %% curly braces.  If you forget this, LaTeX will generate the error
821: %% "Perhaps a missing \item?".
822: %%
823: %% thebibliography produces citations in the text using \bibitem-\cite
824: %% cross-referencing. Each reference is preceded by a
825: %% \bibitem command that defines in curly braces the KEY that corresponds
826: %% to the KEY in the \cite commands (see the first section above).
827: %% Make sure that you provide a unique KEY for every \bibitem or else the
828: %% paper will not LaTeX. The square brackets should contain
829: %% the citation text that LaTeX will insert in
830: %% place of the \cite commands.
831: 
832: %% We have used macros to produce journal name abbreviations.
833: %% AASTeX provides a number of these for the more frequently-cited journals.
834: %% See the Author Guide for a list of them.
835: 
836: %% Note that the style of the \bibitem labels (in []) is slightly
837: %% different from previous examples.  The natbib system solves a host
838: %% of citation expression problems, but it is necessary to clearly
839: %% delimit the year from the author name used in the citation.
840: %% See the natbib documentation for more details and options.
841: 
842: \begin{thebibliography}{}
843: 
844: \bibitem[Agrawal \& Misra (2006)] {am06} Agrawal, V. K. \&
845:    Misra, R., 2006, \apjl, 638, L83
846: \bibitem[Arons (1992)] {a92} Arons, J., 1992, \apj, 388, 561
847: \bibitem[Ballantyne, Turner \& Blaes (2004)] {btb04}
848:    Ballantyne, D. R., Turner, N. J. \& Blaes, O. M., 2004, \apj, 603, 436
849: \bibitem[Ballantyne, Turner \& Young (2005)] {bty05}
850:    Ballantyne, D. R., Turner, N. J., \& Young, A. J., 2005, \apj, 619, 1028
851: \bibitem[Begelman (2001)]{begelman01} Begelman, M. C., 2001, \apj, 551, 897
852: \bibitem[Begelman (2002)]{begelman02} Begelman, M. C., 2002, \apj, 568, L97
853: \bibitem[Begelman (2006)]{begelman06} Begelman, M. C., 2006, \apj, 643, 1065
854: \bibitem[Belloni et al. (2006)] {belloni06} Belloni, T., et al., 2006, 
855:   \apj, 367, 1113
856: \bibitem[B\"ottcher, Jackson \& Liang (2003)] {bjl03}
857:    B\"ottcher, M., Jackson, D. R., \& Liang, E. P., 2003, \apj, 586, 339
858: \bibitem[B\"ottcher \& Liang (2001)] {bl01} B\"ottcher, M. \& Liang, E. P.,
859:    2001, \apj, 552, 248
860: \bibitem[Colbert \& Mushotzky(1999)]{cm99} Colbert, E. J. M., \& Mushotzky,
861:    R. F., 1999, ApJ, 519, 89
862: \bibitem[Colbert \& Ptak(2002)]{cp02} Colbert, E. J. M., \& Ptak, A. F., 2002, 
863:    ApJS, 143, 25
864: \bibitem[Cropper et al.(2004)]{cropper04} Cropper, M., Sorio, R., 
865: Mushotzky, R. F., Wu, K., Markwardt, C. B., \& Pakull, M., 2004, MNRAS, 349, 39
866: \bibitem[Dewangan et al. (2005)] {detal05}
867:    Dewangan, G. C. et al., 2005, \apj, 635, 198
868: \bibitem[Dewangan, Griffiths \& Rao (2006)] {dgr06}
869:    Dewangan, G. C., Griffiths, R. E., \& Rao, A. R., 2006, \apj, 641, L125
870: \bibitem[Dewangan, Titarchuk, \& Griffith(2006)]{dtg06} Dewangan, G.,
871:    Titarchuk, L., \& Griffiths, R. E., 2006, ApJ, 637, L21
872: \bibitem[Fabian et al. (2002)] {fetal02} Fabian, A. C., Ballantyne, 
873:    D. R., Merloni, A., Vaughan, S., Iwasawa, K., \& Boller, Th., \mnras, 
874:    331, L35 
875: \bibitem[Fabbiano (1988)]{fabbiano88} Fabbiano, G., 1988, ApJ, 330, 672
876: \bibitem[Fabbiano et al.(2003)]{fabbiano03} Fabbiano, G., Zezas, A., 
877:    King, A. R., Ponman, T. J., Rots, A., \& Schweizer, F., 2003, ApJ, 584, L5
878: \bibitem[Feng \& Kaaret(2005)]{fk05}Feng, H., \& Kaaret, P., 
879:    2005, ApJ, 633, 1052
880: \bibitem[Finke \& B\"ottcher (2005)] {fb05} Finke, J. D. \&
881:    B\"ottcher, M., 2005, \pasp, 117, 483
882: \bibitem[Gammie (1998)] {g98} Gammie, C. F., 1998, \mnras, 297, 929
883: \bibitem[Georganopoulos, Aharonian \& Kirk (2002)] {gak02}
884:    Georganopoulos, M., Aharonian, F. A., \& Kirk, J. G., 2002, \aap, 388, L25
885: \bibitem[Grevesse, Noels \& Sauval (1996)] {gns96}
886:    Grevesse, N., Noels, A., \& Sauval, A., 1996, in ``Cosmic Abundances''
887:    ASP Conference Series, 99, S. Hold \& G. Sonneborn, eds.
888: \bibitem[Gutierrez (2006)] {g06} Gutierrez, C. M., 2006, \apj, 640, L17
889: \bibitem[Hirose, Krolik, \& Stone (2006)] {hirose06} Hirose, S, 
890:   Krolik, J. H., \& Stone, J. M., 2006, \apj, 640, 901
891: \bibitem[Hui, Krolik \& Hubeny(2005)]{hui05} Hui, Y., Krolik, J. H., \&
892: Hubeny, I., 2005, ApJ, 625, 913
893: \bibitem[Kallman \& Bautista (2001)] {kb01} 
894:    Kallman, T., \& Bautista, M. 2001, \apjs, 133, 221
895: \bibitem[Kalogera et al.(2004)]{kalogera04} Kalogera, V., Henninger, M.,
896: Ivanova, N., \& King, A. R., 2004, ApJ, 603, L41
897: \bibitem[King et al.(2001)]{king01} King, A. R., Davies, M. B., Ward, M. J.,
898: Fabbiano, G., \& Elvis, M., 2001, ApJ, 552, L109
899: \bibitem[Kont et al. (2004)]{ketal04} Kong, A. K. H., Di Stefano, R.,
900:    \& Yuan, F., 2004, \apjl, 617, L49
901: \bibitem[K\"ording, Falcke \& Markoff (2002)]{kfm02} 
902:    K\"ording, E., Falcke, H., \& Markoff, S., 2002, \aap, 382, L13
903: \bibitem[Liu \& Mirabel(2005)]{lm05} Liu, Q. Z., \& Mirabel, I. F., 2005,
904:    A\&A 429, 1125   
905: \bibitem[Madhusudhan et al. (2006)] {metal06} Madhusudhan N., Justham, S.,
906: Paxton, B., Pfahl, E., Podsiadlowski, Ph., \& Rappaport, S., 2006, ApJ, 
907:    640, 918
908: \bibitem[Makishima et al. (2000)] {metal00} Makishima K. et al.,
909:    2000, \apj, 535, 632
910: \bibitem[Markowitz et al.(2003)]{markowitz03} Markowitz, A., et al., 2003,
911:    ApJ, 593, 96
912: \bibitem[Matsumoto et al.(2001)]{matsumoto01} Matsumoto, H., Tsuru, T. G.,
913:    Koyama, K., Awaki, H., Canizares, C. R., Kawai, N., Matsushita, S.,
914: \& Kawabe, R., 2001, ApJ, 547, L25
915: \bibitem[Merloni et al. (2006)] {meretal06} Merloni, A., Malzac, J.,
916:    Fabian, A. C., \& Ross, R. R., 2006, \mnras, 370, 1699 
917: \bibitem[Mizuno et al. (2007)] {mizuno07} Mizuno, T., et al., 
918:    2007, \pasj, 59, 257
919: \bibitem[Nayakshin \& Melia(1998)]{nm98} Nayakshin, S., \& Melia, F.,
920:    1998, \apjs, 114, 269
921: \bibitem[Pakull \& Mirioni (2003)] {pm03} Pakull, M. W., \& Mirioni, L.,
922:   2003, RevMexAA, 15, 197.
923: \bibitem[Pozdnyakov, Sobol \& Sunyaev (1983)] {pss83}
924:    Pozdnyakov, L. A., Sobol, I. M., \& Sunyaev, R. A., 1983, 
925:    Astrophys. Space Phys. Rev., 2, 189
926: \bibitem[Roberts et al. (2006)] {robetal06} 
927:    Roberts, T. P., Kilgard, R. E.,
928:    Warwick, R. S., Goad, M. R., \& Ward, M. J., 2006, \mnras, 
929:    371, 1877 
930: \bibitem[Ross, Fabian \& Ballantyne (2002)] {rfb02} Ross, R. R., 
931:    Fabian, A. C., \& Ballantyne, D. R., 2002, \mnras, 336, 315 
932: \bibitem[Ruszkowski \& Begelman (2003)] {rb03} Ruszkowski, M. \& Begelman, 
933:    M. C., 2003, \apj, 586, 384
934: \bibitem[Saito et al. (2006)] {saito06} Saito, K., Yamaoka, K., 
935:    Fukuyama, M., Miyakawa, T. G., Yoshida, A., \& Homan, J., 
936:    (2007), in ``Sixth Microquasar Workshop: Microquasars and Beyond'', 
937:    astro-ph/0702022
938: \bibitem[Shakura \& Sunyaev(1973)] {ss73} Shakura, N. I., \& Sunyaev, R. A.,
939:    1973, A\&A, 24, 337
940: \bibitem[Stobbart, Roberts \& Wilms (2006)] {srw06} 
941:    Stobbart, A.-M., Roberts, T. P., \& Wilms, J., 2006, \mnras, 368, 397
942: \bibitem[Strohmayer \& Mushotzky (2003)] {sm03} Strohmayer, T. E. \& 
943:    Mushotzky, R. F., 2003, \apj, 586, L61
944: \bibitem[Terashima \& Wilson (2004)] {tw04} Terashima, Y. \&
945:    Wilson, A. S., 2004, \apj, 601, 735
946: \bibitem[Turner et al. (2005)] {tetal05} Turner N. J., Blaes, O. M.,
947:    Socrates, A., Begelman, M. C., \& Davis, S. W., 2005, \apj, 624, 267
948: \bibitem[Winter et al. (2005)] {wetal05} Winter, L. M., Mushotzky, R. F.,
949:    \& Reynolds, C. S., 2005, in ``El Escorial X-ray Universe 2005 meeting'', 
950:    astro-ph/0511493
951: \bibitem[Winter et al. (2006)] {wetal06} Winter, L. M., Mushotzky, 
952:    R. F., Reynolds, C. S., 2006, \apj, 649, 730 
953: 
954: \end{thebibliography}
955: 
956: 
957: \clearpage
958: \begin{deluxetable}{cccccccccccc}
959: \rotate
960: \tabletypesize{\scriptsize}
961: \tablecaption{Simulation and fit parameters.  $\dot{m}$ is the 
962: accretion rate, $\alpha$ is the Shakura-Sunyaev viscosity parameter, 
963: $kT_{in}$ is the inner disk temperature, and $r_{inhom}$ is the
964: radius above which photon bubbles cannot exist ($l<1$).  
965: Results of our simulations are:
966: total luminosity over the Eddington luminosity is given 
967: ($L_{tot} / L_{Edd}$); the 0.2---10 keV luminosity over the 
968: Eddington luminosity ($L_X / L{Edd}$);
969: the fitted photon index, $\Gamma$; 
970: the high energy exponential
971: cutoff; and the fraction of the 
972: total luminosity in the MCDBB component($f_{BB}$), in the Comptonization 
973: component ($f_{Compt}$), and the reflection component ($f_{refl}$).  
974: Note that these do not all add up to unity due to rounding.
975: }
976: \tablewidth{0pt}
977: \tablehead{ 
978: \colhead{ Sim. No. } &
979: \colhead{ $\dot{m}$ } & 
980: \colhead{ $\alpha$ } &
981: \colhead{ $kT_{in}$ [keV] } &
982: \colhead{ $r_{inhom}$ } &
983: \colhead{ $L_{tot} / L_{Edd}$ } & 
984: \colhead{ $L_X / L{Edd}$ } &
985: \colhead{ $\Gamma$ } &
986: \colhead{ $E_{cutoff}$ [keV] } &
987: \colhead{ $f_{BB}$ } &
988: \colhead{ $f_{Compt}$ } &
989: \colhead{ $f_{refl}$ }
990: } 
991: \startdata
992: 1  & 200 & 0.01 & 1.4 &  280 & 2.8  & 1.3 &  2.18 & 460 & 0.22 & 0.73 & 0.05 \\
993: 2  & 400 & 0.01 & 1.6 &  472 & 4.4  & 2.5 &  2.42 & 440 & 0.26 & 0.63 & 0.11 \\
994: 3  & 900 & 0.01 & 2.0 &  860 & 6.9  & 4.7 &  2.42 & 290 & 0.56 & 0.40 & 0.05 \\
995: 4  & 200 & 0.1  & 1.4 &  482 & 5.0  & 1.7 &  1.96 & 280 & 0.17 & 0.82 & 0.01 \\
996: 5  & 400 & 0.1  & 1.6 &  804 & 6.5  & 2.9 &  2.12 & 220 & 0.27 & 0.74 & 0.01 \\
997: 6  & 900 & 0.1  & 2.0 & 1456 & 9.4  & 5.3 &  2.24 & 140 & 0.39 & 0.59 & 0.02 \\
998: 7  & 200 & 0.5  & 1.4 &  698 & 5.8  & 1.8 &  1.75 & 190 & 0.14 & 0.85 & 0.01 \\
999: 8  & 400 & 0.5  & 1.6 & 1164 & 8.5  & 3.1 &  1.78 & 160 & 0.17 & 0.79 & 0.05 \\
1000: 9  & 900 & 0.5  & 2.0 & 2100 & 11.9 & 5.6 &  1.95 & 160 & 0.31 & 0.68 & 0.02 \\
1001: \enddata
1002: \label{parameters}
1003: \end{deluxetable}
1004: \clearpage
1005: 
1006: %% Use the figure environment and \plotone or \plottwo to include
1007: %% figures and captions in your electronic submission.
1008: %% To embed the sample graphics in
1009: %% the file, uncomment the \plotone, \plottwo, and
1010: %% \includegraphics commands
1011: %%
1012: %% If you need a layout that cannot be achieved with \plotone or
1013: %% \plottwo, you can invoke the graphicx package directly with the
1014: %% \includegraphics command or use \plotfiddle. For more information,
1015: %% please see the tutorial on "Using Electronic Art with AASTeX" in the
1016: %% documentation section at the AASTeX Web site,
1017: %% http://www.journals.uchicago.edu/AAS/AASTeX.
1018: %%
1019: %% The examples below also include sample markup for submission of
1020: %% supplemental electronic materials. As always, be sure to check
1021: %% the instructions to authors for the journal you are submitting to
1022: %% for specific submissions guidelines as they vary from
1023: %% journal to journal.
1024: 
1025: \begin{figure}
1026: \epsscale{1.0}
1027: \plotone{f1_color.eps}%{f1.eps}
1028: %\plotone{geometry2.eps}
1029: \caption{The simulation geometry.  Each simulation is of one part of
1030: the disk.  The LDR is divided into 40 zones.  Afterwords, simulations 
1031: of different parts of the disk are averaged, weighted by area.
1032: See the electronic edition of the Journal for a color version of this figure.
1033: \label{geometry}}
1034: \end{figure}
1035: 
1036: \clearpage
1037: 
1038: \begin{figure}
1039: \epsscale{1.0}
1040: \plotone{f2_color.eps}%{f2.eps}
1041: %\plotone{specradii.eps}
1042: \caption{Spectra at various radii and the total spectrum for 
1043: Simulation 2.  See the electronic edition of the Journal for a color 
1044: version of this figure.  }
1045: \label{specradii}
1046: \end{figure}
1047: 
1048: \clearpage
1049: 
1050: \begin{figure}
1051: \epsscale{1.0}
1052: \plotone{f3_color.eps}%{f3.eps}
1053: %\plotone{paramradius1.eps}
1054: \caption{The parameters $\delta$ (a), $\xi$ (b), $\tau$ (c), 
1055: and $l$ (d) as a 
1056: function of radius, for $\dot{m}=200$ (solid black), $\dot{m}=400$ 
1057: (dotted red), 
1058: and $\dot{m}=900$ (dashed green).  These simulations have $\alpha=0.01$.  
1059: See the electronic edition of the Journal for a color version of this figure.}
1060: \label{paramradius1}
1061: \end{figure}
1062: 
1063: \clearpage
1064: 
1065: \begin{figure}
1066: \epsscale{1.0}
1067: \plotone{f4_color.eps}%{f4.eps}
1068: %\plotone{paramradius2.eps}
1069: \caption{Same as Fig. \ref{paramradius1} except for $\alpha=0.1$.
1070: See the electronic edition of the Journal for a color version of this figure.}
1071: \label{paramradius2}
1072: \end{figure}
1073: 
1074: \clearpage
1075: 
1076: \begin{figure}
1077: \epsscale{1.0}
1078: \plotone{f5_color.eps}%{f5.eps
1079: %\plotone{paramradius3.eps}
1080: \caption{Same as Fig. \ref{paramradius1} except for $\alpha=0.5$.
1081: See the electronic edition of the Journal for a color version of this figure.}
1082: \label{paramradius3}
1083: \end{figure}
1084: 
1085: \clearpage
1086: 
1087: \begin{figure}
1088: \epsscale{1.0}
1089: \plotone{f6_color.eps}%{f6.eps}
1090: %\plotone{spectot.eps}
1091: \caption{Total spectra for (a) $\alpha=0.01$, (b) $\alpha=0.1$, and
1092: (c) $\alpha=0.5$.  For all graphs we plot $\dot{m}=200$ (solid black), 
1093: $\dot{m}=400$ (dotted red), and $\dot{m}=900$ (dashed green).  
1094: See the electronic edition of the Journal for a color version of this 
1095: figure.  }
1096: \label{spectot}
1097: \end{figure}
1098: 
1099: \clearpage
1100: 
1101: \begin{figure}
1102: \epsscale{1.0}
1103: \plotone{f7_color.eps}%{f7.eps}
1104: %\plotone{reflect.eps}
1105: \caption{The spectra broken into components for $\alpha=0.01$ and 
1106: (a) $\dot{m}=200$, (b) $\dot{m}=400$, and (c) $\dot{m}=900$.  
1107: The solid black line is the reflected component, the dotted red line is the 
1108: MCDBB, the dashed green line is the Comptonization 
1109: component, and the long dashed blue line is the total spectrum. 
1110: See the electronic edition of the Journal for a color version of this 
1111: figure.  }
1112: \label{reflect}
1113: \end{figure}
1114: 
1115: \clearpage
1116: 
1117: 
1118: \end{document}
1119: %%
1120: %% End of file `sample.tex'.
1121: