1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[manuscript]{aastex}
3: %\documentstyle[aasms4]{article}
4: %\documentstyle[12pt,aaspp4]{article}
5:
6:
7:
8: \newcommand{\Rs}{$R_{\rm s}$}
9: \newcommand{\HR}{H/R}
10: \newcommand{\dl}{$dL/dR$}
11: \newcommand{\astar}{$a_*$}
12: \newcommand{\al}{$\alpha$}
13: \newcommand{\cs}{$\c_{\rm s}^2$}
14: \newcommand{\omk}{$\Omega_{\rm K}$}
15: \newcommand{\mdot}{$\dot{M}$}
16: \newcommand{\Tmax}{$T_{\rm max}$}
17: \newcommand{\Rg}{$R_{\rm g}$}
18: \def\sles{\lower2pt\hbox{$\buildrel {\scriptstyle <}
19: \over {\scriptstyle\sim}$}}
20: %insert forced spacing before and after \sles, \sgreat within $--$
21:
22: \def\sgreat{\lower2pt\hbox{$\buildrel {\scriptstyle >}
23: \over {\scriptstyle\sim}$}}
24: %needs forced spacing, as above
25:
26:
27:
28: \begin{document}
29:
30: \title{VISCOUS TORQUE AND DISSIPATION IN THE INNER REGIONS OF A THIN
31: ACCRETION DISK: IMPLICATIONS FOR MEASURING BLACK HOLE SPIN}
32:
33: \author{Rebecca Shafee\altaffilmark{1},
34: Ramesh Narayan\altaffilmark{2},
35: Jeffrey E. McClintock\altaffilmark{2}}
36:
37: \altaffiltext{1}{Harvard University, Department of Physics,
38: 17 Oxford Street, Cambridge, MA 02138}
39: \altaffiltext{2}{Harvard-Smithsonian Center for Astrophysics, 60 Garden
40: Street, Cambridge, MA 02138}
41:
42:
43:
44:
45: \begin{abstract}
46:
47: We consider a simple Newtonian model of a steady accretion disk around
48: a black hole. The model is based on height-integrated hydrodynamic
49: equations, $\alpha$-viscosity, and a pseudo-Newtonian potential which
50: results in an innermost stable circular orbit (ISCO) that closely
51: approximates the one predicted by GR. We find that, as the disk
52: thickness $H/R$ or the value of $\alpha$ increases, the hydrodynamic
53: model exhibits increasing deviations from the standard thin disk model
54: of Shakura \& Sunyaev. The latter is an analytical model in which the
55: viscous torque is assumed to vanish at the ISCO. We consider the
56: implications of the results for attempts to estimate black hole spin
57: by using the standard disk model to fit continuum spectra of black
58: hole accretion disks. We find that the error in the spin estimate is
59: quite modest so long as $H/R \leq 0.1$ and $\alpha \leq 0.2$. At
60: worst the error in the estimated value of the spin parameter is 0.1
61: for a non-spinning black hole; the error is much less for a rapidly
62: spinning hole. We also consider the density and disk thickness
63: contrast between the gas in the disk and that inside the ISCO. The
64: contrast needs to be large if black hole spin is to be successfully
65: estimated by fitting the relativistically-broadened X-ray line profile
66: of fluorescent iron emission from reflection off an accretion disk.
67: In our hydrodynamic models, the contrast in density and thickness is
68: low when $H/R\ \sgreat\ 0.1$, suggesting that the iron line technique
69: may be most reliable in extremely thin disks. We caution that these
70: results have been obtained with a viscous hydrodynamic model. While
71: our results are likely to be qualitatively correct, quantitative
72: estimates of, e.g., the magnitude of the error in the spin estimate,
73: need to be confirmed with MHD simulations of radiatively cooled thin
74: disks.
75:
76:
77: \end{abstract}
78:
79: \keywords{X-ray: stars --- binaries: close --- accretion, accretion
80: disks --- black hole physics }
81:
82: \normalsize
83:
84: \section{INTRODUCTION}
85:
86: Recently, we reported spin estimates of three black holes (BHs) in
87: Galactic X-ray binaries (Shafee et al. 2006; McClintock et al. 2006;
88: hereafter S06, M06). The results were obtained by fitting the soft
89: X-ray continuum spectra of these systems in the thermal state
90: (Remillard \& McClintock 2006) to a general relativistic, multicolor
91: blackbody, thin disk model ($Kerrbb$, Li et al. 2005), which includes
92: the effect of spectral hardening (Davis et al. 2005). In this method,
93: which was pioneered by Zhang, Cui \& Chen (1997), we assume a
94: razor-thin disk that terminates at the innermost stable circular orbit
95: (ISCO). In addition, we assume that the viscous torque vanishes at the
96: ISCO and that there is no energy dissipation or angular momentum loss
97: inside the ISCO. These are standard assumptions in the theory of
98: accretion disks (e.g., Shakura \& Sunyaev 1973; Frank, King \& Raine
99: 2002), and correspond to what we refer to in this paper as the
100: ``standard disk model.'' However, there has been debate in recent
101: times as to the validity of the assumptions.
102:
103: The stress responsible for angular momentum transport in a thin
104: accretion disk is likely to be magnetic (Balbus \& Hawley 1991). If
105: this is the case, an argument could be made for a non-zero stress at
106: the ISCO, coupled with considerable dissipation near and inside the
107: ISCO (Krolik 1999; Gammie 1999). These effects could cause
108: important deviations from the standard disk model (Krolik \& Hawley
109: 2002), perhaps invalidating our spin determinations.
110:
111: Afshordi \& Paczy\'nski (2003), following earlier work by
112: Abramowicz \& Kato (1989) and Paczy\'nski (2000), suggested that the
113: torque at the ISCO increases with increasing disk thickness.
114: Motivated by their work, we argued in M06 that deviations from the
115: standard disk model are likely to be serious only for thick disks. We
116: thus restricted our attention to relatively thin disks with
117: height-to-radius ratios of $\HR<0.1$. The present paper is an attempt
118: to verify whether or not such thin disks do indeed behave like the
119: standard disk model.
120:
121: In addition to the debate over the validity of using the standard
122: disk theory to model the continuum spectra of realistic disks,
123: another relevant issue in attempting to estimate BH spin is the
124: relative merit of the continuum fitting method compared to fitting
125: the relativistically-broadened fluorescent iron line in the X-ray
126: spectrum. Both methods have been proposed as a means of estimating
127: BH spins, and it is of interest to understand how well the
128: assumptions of each are satisfied by real disks. The models
129: currently used by the iron line method assume that the line
130: emissivity peaks at the ISCO, drops abruptly to zero inside the
131: ISCO, and decreases steeply as a broken power-law outside the ISCO
132: (e.g., Brenneman \& Reynolds 2006, hereafter BR06). This requires,
133: among other things, a significant drop in matter density (Fabian
134: 2007) or disk thickness (Nayakshin et al. 2000, 2002) inside the
135: ISCO. A second motivation for the present paper is therefore to
136: check the validity of the assumed line emissivity profile.
137:
138: Our analysis is based on a non-relativistic hydrodynamic model of an
139: accretion disk. We present global numerical solutions of the
140: differential equations governing the fluid flow, assuming that the
141: accretion disk is steady, axisymmetric and in hydrostatic
142: equilibrium in the vertical direction, and using a pseudo-Newtonian
143: model for the gravitational potential. We do not include magnetic
144: fields explicitly, but assume an effective viscosity described by
145: the \al\ prescription (Shakura \& Sunyaev 1973). We also assume an
146: adiabatic index $\gamma=1.5$, which corresponds to approximate
147: equipartition between gas and magnetic pressure (Quataert \& Narayan
148: 2000).
149:
150: Our primary interest is in accretion disks in the rigorously defined
151: thermal state (see Table 2 in Remillard \& McClintock 2006) with $\HR
152: <0.1$, as these are the systems of most interest for our work on BH
153: spin (M06). Since the value of the viscosity parameter \al\ for such
154: disks is a matter of debate, we try different constant values: \al~=
155: 0.01, 0.1, 0.2. We also consider a variable-\al\ prescription (eq. 22)
156: inspired by the MHD simulations of Hawley \& Krolik (2002, hereafter
157: HK02). For non-spinning BHs, we use the pseudo-Newtonian potential of
158: Paczy\'nski \& Wiita (1980; hereafter PW80), and for spinning black
159: holes we use the pseudo-Kerr model of Mukhopadhyay (2002). The
160: numerical framework for our calculations is similar to that used by
161: Narayan, Kato \& Honma (1997), viz., we use a relaxation method to
162: solve the equations from the sonic radius \Rs\ to the outer edge of
163: the disk ($\sim 10^5$ \Rs), and we then integrate inward from \Rs\ to
164: the event horizon.
165:
166: The paper is organized as follows. We discuss in \S2 the theory and
167: computational method. We then discuss in \S3 our numerical disk
168: solutions, focusing on the magnitude of the stress at the ISCO, the
169: amount of viscous dissipation near and inside the ISCO, and the
170: density and disk thickness contrast across the ISCO.
171: We then compute in \S4 the
172: emitted spectra of our numerical disks for different values of $\HR$
173: and \al\ and investigate the error we make when we estimate the spin
174: of a BH via the continuum fitting method assuming the standard disk
175: model. We conclude in \S5 with a discussion.
176:
177:
178: \section
179: {THE MODEL}
180:
181: \subsection{Gravity}
182:
183: In order to focus our attention on the key physics of the problem,
184: and to avoid being distracted by technical details, we consider a
185: simple viscous hydrodynamic accretion disk in a Newtonian
186: gravitational potential. Since the presence of an ISCO is essential
187: for our analysis, we simulate relativistic gravity in this Newtonian
188: model by means of a modified gravitational potential. For a
189: non-spinning BH, we make use of the PW80 potential:
190: \begin{equation}
191: \Phi=-\frac{G M}{R-2 R_{\rm g}}\; ,
192: \end{equation}
193: where $M$ is the BH mass, $G$ the gravitational constant and $R_{\rm
194: g}=GM/c^2$. The Keplerian angular velocity \omk\ at a radius $R$
195: from the BH is
196: \begin{equation}
197: \Omega_{\rm K}=\frac{ (G M)^{1/2}}{(R-2 R_{\rm g}) R^{1/2}}\; .
198: \end{equation}
199: In the case of a spinning BH we use the pseudo-Kerr model of
200: Mukhopadhyay (2002) in which the gravitational acceleration of a
201: test particle in a Keplerian orbit at a distance $R$ from the BH is
202: \begin{equation}
203: F=- \nabla \Phi = \frac{c^4}{G M}\frac{(r^2- 2 a_* \sqrt{r} +
204: a_*^2)^2}{r^3(\sqrt{r}(r-2)+a_*)^2}\; ,
205: \end{equation}
206: where $r=R/R_{\rm g}$, $a_*=a/M=J/(GM^2/c)$ is the dimensionless
207: spin of the BH, and $-1< a_*<1$. The Keplerian angular velocity at
208: radius $R$ is then
209: \begin{equation}
210: \Omega_{\rm K}=\frac{c^3}{G M}\frac{(r^2 - 2 a_* \sqrt{r} +a_*^2)}{r^2
211: (\sqrt{r} (r-2)+a_*)} \; .
212: \end{equation}
213:
214: \subsection{Hydrodynamics}
215:
216:
217: We assume a steady axisymmetric disk in hydrostatic equilibrium in the
218: vertical direction. In the equations that follow, which have a long
219: history in accretion disk theory (e.g., Paczy\'nski \&
220: Bisnovatyi-Kogan 1981; Muchotrzeb \& Paczy\'nski 1982; Kato, Honma \&
221: Matsumoto 1988; Abramowicz et al. 1988; Popham \& Narayan 1991;
222: Narayan \& Popham 1993; Chen \& Taam 1993; Narayan et al. 1997; Chen,
223: Abramowicz \& Lasota 1997), we denote density, sound speed, radial
224: velocity, angular velocity, Keplerian angular velocity, and vertical
225: half-thickness by $\rho$, $c_{\rm s}$, $v_{\rm R}$, $\Omega$, \omk,
226: and $H$, respectively. All these parameters are taken to be functions
227: of the cylindrical radius $R$ only. Because of the assumption of
228: steady state, the Lagrangian time derivative $D/Dt= \partial/\partial
229: t +\mathbf{v }\cdot \mathbf{\nabla}$ becomes $D/Dt=v_{\rm R}d/dR$.
230: After vertical and then radial integration the continuity equation
231: takes the form:
232: \begin{equation}
233: 4 \pi \rho v_{\rm R} R H=-\dot{M}={\rm constant}\; ,
234: \end{equation}
235: where $H= c_{\rm s}/\Omega_{\rm K}$. The momentum equation is
236: \begin{equation}
237: \rho (\mathbf{v \cdot \nabla}) \mathbf{v}=-\mathbf{\nabla} P - \rho
238: \mathbf{\nabla} \Phi
239: +\rho \Omega^2 \mathbf{R} +\rho \mathbf{\nabla} \cdot
240: \mathit{\sigma}\; ,
241: \end{equation}
242: where ${\it \sigma}$ is the stress tensor. We assume that the only
243: non-zero component of ${\it \sigma}$ is ${\it \sigma_{\rm {R
244: \Phi}}}= - \alpha P$ (\al\ prescription, Shakura \& Sunyaev 1973),
245: where $P$ is the total pressure and we write $P= \rho c_{\rm s^2}$.
246: The radial component of the momentum equation gives
247: \begin{equation}
248: v_{\rm R} \frac{d v_{\rm R}}{d R}=-(\Omega_{\rm K}^2 -
249: \Omega^2)R -\frac{1}{\rho} \frac{d}{d R}(\rho c_{\rm s}^2) \; ,
250: \end{equation}
251: and conservation of angular momentum gives
252: \begin{equation}
253: \frac{\rho v_{\rm R}}{ R} \frac{d}{d R} (\Omega R^2)=
254: \frac{1}{R^2 H} \frac{d(R^2 H \mathit{\sigma_{\rm {R
255: \Phi}}})}{d R}\; .
256: \end{equation}
257: The latter equation can be integrated to obtain
258: \begin{equation}
259: \Omega R^2 -j =-\frac{\alpha c_{\rm s}^2 R}{v_{\rm R}}\; ,
260: \end{equation}
261: where $\Omega R^2$ is the specific angular momentum of the gas at
262: radius $R$ and $j$ is an integration constant. We can interpret $j$ as
263: the specific angular momentum of the accreting gas at the radius where
264: the stress goes to zero.
265:
266: Lastly, we write the energy conservation equation in terms of the
267: Lagrangian derivative of the specific entropy,
268: \begin{equation}
269: \rho T \frac{D s}{Dt}= q^+- q^-= f q^+.
270: \end{equation}
271: Here $s$ is the specific entropy per unit mass, and $q^+$ and $q^-$
272: are the volume rate of heating and cooling of the gas, respectively.
273: Following Narayan et al. (1997) we take the cooling rate to be a
274: factor $(1-f)$ of the heating rate. Narayan et al. used $f=1$ because
275: they were modeling advection-dominated accretion flows. Since we are
276: interested primarily in thin disks, we use small values of $f$, i.e.,
277: substantial cooling, and we tune the value of $f$ to achieve the
278: desired disk thickness (eq. 19). The heating of the gas is due to
279: viscous dissipation, which gives $ q^+= \nu \sigma R
280: d\Omega/dR$. Using the relationship $\epsilon= P/(\gamma-1)$, where
281: $\epsilon$ is the thermal energy per unit volume and $\gamma$ is the
282: adiabatic index (we use $\gamma=1.5$), we can write
283: \begin{equation}
284: \rho T \frac{D s}{Dt}= \frac{\rho v_{\rm R}}{\gamma-1}\frac{d c_{\rm
285: s}^2}{dt}
286: - c_{\rm s}^2 v_{\rm R} \frac{d \rho}{d R}\; .
287: \end{equation}
288: Thus, the energy equation takes the form
289: \begin{equation}
290: \frac{\rho v_{\rm R}}{\gamma-1}\frac{d c_{\rm s}^2}{d R}- c_{\rm s}^2 v_{\rm R}
291: \frac{d \rho}{d R}=-f\alpha \rho c_{\rm s}^2 R \frac{d \Omega}{dR}\; .
292: \end{equation}
293:
294:
295: \subsection{Boundary Conditions and Numerical Method}
296:
297: We use a relaxation method to obtain numerical solutions of the above
298: differential equations. In the computations, we define $x=R/R_{\rm
299: s}$ as the spatial variable and covered the region $x=1$ to $10^5$
300: using 1000 grid points. The grid has a non-uniform spacing, with more
301: grid points near the inner boundary $x=1$. In solving the equations,
302: we set $- 4 \pi \rho v_{\rm R} R H = \dot{M}=1$, and in order to
303: simplify the equations we substitute for $\rho$ using equation
304: (5). Thus we are left with three unknown functions of $R$: $v_{\rm
305: R}(R)$, $c_{\rm s}^2(R)$, and $\Omega(R)$. In addition, we have two
306: unknown constants, $j$ and $R_s$, which we treat as eigenvalues. To
307: solve for these quantities, we use equations (7), (9), and (12),
308: supplemented with five boundary conditions.
309:
310: Narayan et al. (1997) showed that solutions of the disk model
311: described in \S2.2 tend to be nearly self-similar over a wide range of
312: radius. Assuming self-similarity (following Narayan \& Yi 1994), we
313: can obtain the following analytic solution of the equations (the
314: subscript ``SS'' refers to self-similar):
315: \begin{equation}
316: c^2_{\rm s,SS}(R)= c_0^2 \frac{G M}{R}\; , \; c_0^2=\frac{2}{5+ 2
317: \epsilon'+ \alpha^2 / \epsilon'}\; , \;
318: \epsilon'=\frac{5/3-\gamma}{f(\gamma-1)}\; ,
319: \end{equation}
320: \begin{equation}
321: v_{\rm R,SS} (R)= v_0 \sqrt{\frac{G M}{R}}\; ,\;
322: v_0=-\alpha \sqrt{\frac{c_0^2}{\epsilon'}}\; ,
323: \end{equation}
324: \begin{equation}
325: \Omega_{\rm SS}(R)= \Omega_0 \Omega_K \; , \; \Omega_0=\sqrt{\frac{2
326: \epsilon'}{5+ 2 \epsilon'+ \alpha^2/\epsilon'}}\; .
327: \end{equation}
328: We use this self-similar solution to set boundary conditions at the
329: outer boundary $R_{\rm
330: out}$:
331: \begin{equation}
332: v_{\rm R}(R_{\rm out})=v_0\sqrt{\frac{G M}{R_{\rm out}}}\; ,
333: \end{equation}
334: \begin{equation}
335: c^2_{\rm s}(R_{\rm out})=c_0^2 \frac{G M}{R_{\rm out}}\; ,
336: \end{equation}
337: \begin{equation}
338: \Omega(R_{\rm out})=\Omega_0 \Omega_{\rm K}\;.
339: \end{equation}
340: From the above relations it can be shown that, at the outer
341: boundary, the vertical scale-height $H$ satisfies
342: \begin{equation}
343: \frac{H}{R}=\sqrt{\frac{2}{5+2 \epsilon'+ \alpha^2/\epsilon'}} \; .
344: \end{equation}
345: Therefore, for a given value of $\gamma$, we can vary the disk
346: thickness $\HR$ by changing $f$. For $\gamma=1.5$, $f$= 0.000035 and
347: 0.0035 give $H/R=0.01$ and 0.1, respectively. Once set at the outer
348: edge, the value of $H/R$ remains constant over most of the disk,
349: becoming smaller only near and inside the ISCO. Note that, for the
350: thin disk models that we consider in this paper which have $H/R \leq
351: 0.1$, the advection parameter $f$ is very much less than unity. This
352: means that radiative cooling (which is $\propto 1-f$) dominates by a
353: huge factor over energy advection ($\propto f$). We briefly
354: discuss thicker advection-dominated solutions in \S5.
355:
356: The inner boundary is at the sonic radius, $R=R_{\rm s}$, which is a
357: singular point of the differential equations. Following standard
358: methods, we obtain the following regularity conditions at $R_{\rm s}$:
359: \begin{equation}
360: v_{\rm R}^2-\frac{2 \gamma}{\gamma+1}c_{\rm s}^2=0\; ,
361: \end{equation}
362: \begin{equation}
363: (\Omega_{\rm K}^2- \Omega^2)R- c_{\rm s}^2 \frac{2
364: \gamma}{\gamma+1}\left(\frac{1}{R}-\frac{d
365: \ln(\Omega_{\rm K})}{dR}\right)-c_{\rm s}^2
366: \frac{\gamma-1}{\gamma+1}\frac{f
367: \alpha R}{v_{\rm R}}\frac{d \Omega}{dR}=0.
368: \end{equation}
369: Equations (16)--(18), (20)--(21) provide the five boundary
370: conditions we need to find a unique solution. Once we have obtained
371: the solution between $R=R_{\rm s}$ and $R= R_{\rm out}$ via the
372: relaxation method, we use the solution at $R=R_{\rm s}$ as initial
373: conditions and integrate the equations from \Rs\ down close to the
374: BH event horizon.
375:
376: We should emphasize that we do not set any boundary condition at
377: the ISCO. Instead, we apply the boundary conditions at the sonic
378: radius, whose position is computed self-consistently for each
379: solution. Further, even at the sonic radius, the viscous torque is
380: not set to zero --- the torque is computed self-consistently and is
381: allowed to continue smoothly inside the ISCO. The numerical solutions
382: we obtain are thus superior to the standard disk model and can be used
383: to check the validity of the latter. In particular, we can estimate
384: what error one makes in the standard disk model as a result of the
385: zero-torque boundary condition.
386:
387:
388: \section{RESULTS}
389:
390: \subsection{Numerical Solutions}
391:
392: Figure 1 shows model results for a non-spinning BH. We consider two disk
393: thicknesses: $\HR=0.01$ (solid lines), and $\HR=0.1$ (dotted lines). In
394: all four panels the vertical line shows the position of the ISCO ($R=6
395: R_{\rm g}$). We use $G=M=c=1$, so that the unit of velocity and time are
396: c and $G M/c^3$, respectively, and set $\dot{M}=1$. Most of our models
397: correspond to a constant value of \al. However, we also consider a model
398: in which \al\ varies as a function of $R$,
399: \begin{equation}
400: \alpha= \frac{16.8}{(R/R_{\rm g})^3} +0.1 \; ,
401: \end{equation}
402: which closely reproduces the effective profile of \al\ found by HK02
403: (see their Fig. 4). We refer to this as the ``variable-$\alpha$
404: model.''
405:
406: Figure $1a$ shows the variation of the sound speed squared $c_{\rm
407: s}^2$ as a function of radius $R$. For a given thickness, the
408: different \al\ models overlap at large radii and are only
409: distinguishable in the inner region of the disk. Here and in the
410: figures that follow, the magenta, blue, red and green lines refer to
411: the \al~= 0.01, 0.1, 0.2 and variable-\al~models, respectively.
412: Figure $1b$ shows the radial infall velocity {$v_{\rm R}$} of the
413: accreting gas. We see that, between the ISCO and the event horizon,
414: {$v_{\rm R}$} increases rapidly regardless of the value of \al. The
415: variable-\al\ model almost completely overlaps with the \al = 0.1
416: model even at large radii. Figure $1c$ shows the angular velocity
417: $\Omega$ and Keplerian angular velocity \omk. The profiles of $\Omega$
418: for the different values of \al\ and $H/R$ are not distinct and are
419: represented by the single dotted line. The solid red line corresponds
420: to the Keplerian angular velocity. Note that the gas orbits in a
421: nearly Keplerian fashion until it reaches the ISCO. Thereafter, the
422: hydrodynamic forces maintain an orbital motion that becomes
423: increasingly sub-Keplerian as the gas approaches the event
424: horizon. Figure $1d$ shows the gas density $\rho$ as a function of
425: radius. As in the case of the sound speed (Fig. $1a$), the density
426: reaches a maximum outside the ISCO and then decreases rapidly near the
427: event horizon. In this plot, too, the variable-\al\ model coincides
428: with the \al~= 0.1 model.
429:
430: Figure 2 is in the same format as Figure 1 and presents our results
431: for a spinning BH with $a_*=0.95$. The principal difference from the
432: previous figure is that the ISCO (vertical dashed line) is now located
433: at $R=1.937R_{\rm g}$. We consider the same values of $\alpha$ and
434: $H/R$ as in Figure 1, but there is no variable-$\alpha$ model in this
435: case because HK02 considered only a non-spinning BH.
436:
437: \subsection{Matter Density, Disk Thickness and the Iron Line Method}
438:
439: Before presenting our main results in the following subsections, we
440: briefly consider the implications of our models for the determination
441: of spin via the iron line method. The source geometry and illumination
442: law for producing the fluorescence iron line are probably the largest
443: uncertainties in the line fitting method (Reynolds \& Begelman
444: 1997). If we assume the steepest law that is suggested by Reynolds \&
445: Begelman (1997), then the irradiating flux $F_{\rm X} \sim R^{-3}$.
446: Let us write the emissivity function in the form $ f_{\rm Fe} F_{\rm
447: X}$, where $f_{\rm Fe}$ is an efficiency factor. In this section we
448: investigate if the existing models of $f_{\rm Fe}$ in the literature
449: agree with our hydrostatic models.
450:
451: The currently favored iron line models (BR06) assume that the iron
452: line emission is restricted between $R_{\rm ISCO}$ and an outer radius
453: $R_{\rm out}$ and that, within this region, the line profile is fitted
454: by a broken power law. BR06 find that the emissivity varies as $\sim
455: R^{-6}$ between the break radius $R_{\rm br}$ and $R_{\rm ISCO}$, and
456: as $\sim R^{-3}$ between $R_{\rm br}$ and $R_{\rm out}$. For $F_{\rm
457: X} \sim R^{-3}$, this implies the following form for the efficiency
458: function:
459: \begin{equation}
460: \begin{array}{lll}
461: f_{\rm Fe(BR06)} & = 0 \; , & {\rm if }\; \: R < R_{\rm ISCO} \; ,\\
462: & = 1/R^3 \; , & {\rm if}\; \:R_{\rm ISCO}
463: \le R \le 3\:R_{\rm ISCO} \; , \\
464: & = {\rm constant}\; , & {\rm if }\;
465: \: 3\:R_{\rm ISCO} < R \; .
466: \end{array}
467: \end{equation}
468: Below we discuss the two main theories regarding the physical
469: parameters that might affect the emissivity profile.
470:
471: Constant density models (Ross \& Fabian 2003; \.{Z}ycki et al. 1994;
472: Ross, Fabian \& Young 1999) predict that the line emissivity is
473: dependent on the ionization parameter, which is proportional to
474: $F_{\rm X}/\rho$, where $\rho$ is the gas density and $F_{\rm X}$ is
475: the illuminating flux. It is argued that the gas density drops to very
476: low values inside the ISCO. As a result, the region inside the ISCO
477: has a very high ionization parameter, which in turn produces
478: negligible iron line emission (Reynolds \& Begelman 1997; Young et
479: al. 1998; Fabian 2006). In this case, one would expect $f_{\rm Fe}$
480: to be inversely related to ionization, i.e., $f_{\rm Fe}$ should be a
481: function of $\rho(R)/F_X(R) \propto \rho(R) R^3$ . More detailed
482: calculations that solve for the vertical structure of the disk under
483: hydrostatic equilibrium (e.g. Nayakshin et al. 2000, 2002) suggest
484: that the line emission depends on a ``gravity factor'' $\sim
485: (H/R^3)F_{\rm X}$. If that is the case, then for $F_{\rm X} \sim
486: R^{-3}$, one expects the efficiency function $f_{\rm Fe}$ to be
487: proportional to $H$.
488:
489: In Figure 3, we compare the BR06 efficiency function $f_{\rm
490: Fe(BR06)}$ (eq. 23) to those suggested by our hydrostatic models,
491: in the context of the constant density and gravity theories mentioned
492: above. Figure $3a$ shows $\rho(R) R^3$ as a function of radius. The
493: line types/colors for the various models are the same as those defined
494: in Figure 1. Superimposed on our density profiles is a thick
495: short-dashed black line that represents $f_{\rm Fe(BR06)}$. For both
496: $\HR=0.01$ (solid lines) and $\HR=0.1$ (dotted lines), and all values
497: of \al, we note that $\rho(R) R^3$ is an increasing function of
498: radius, implying that $f_{\rm Fe}$ should also increase with
499: increasing radius. There is no apparent reason why $f_{\rm Fe}$ should
500: increase so steeply near the ISCO, or decrease at large radii, as
501: suggested by $f_{\rm Fe(BR06)}$.
502:
503:
504: Figure $3b$ shows a similar plot for a rapidly spinning BH with
505: $a_*=0.95$. We notice the same trends as in Figure $3a$. In this case, we also
506: notice that (especially for $\HR=0.1$), instead of becoming negligible
507: at the ISCO, $\rho(R) R^3$ decreases gradually as one passes the ISCO
508: and moves closer to the event horizon. Therefore, one does not expect
509: $f_{\rm Fe}$ to drop abruptly to zero at the ISCO.
510:
511: In Figures $3c$ and $3d$, we consider the disk thickness in the inner
512: region. In our models, the disk has a more or less constant thickness
513: specified by $H/R$ outside $\sim 100 R_{\rm g}$, and we vary this
514: ``outer thickness'' by changing the value of $f$ (eq. 19). However,
515: in the inner region, the disk gets thinner. Figure $3c$ shows $H$ as
516: a function of $R$ for a non-spinning BH. The top panel shows a disk
517: with outer thickness of 0.01 and the bottom one shows a disk with
518: outer thickness of 0.1 for the choices of $\alpha$ specified in \S3.1.
519: In the thinner case, there is an abrupt drop in $H$, which would
520: likely quench the iron emission from inside the ISCO. For the thicker
521: case, however, the value of $H$ decreases gradually and remains
522: significant far inside the ISCO at $3R_{\rm g}$. Thus, these models
523: indicate that the region within the ISCO may contribute a significant
524: fraction of the total iron line emission and, also that it is
525: difficult to justify the steeply falling form of $f_{\rm
526: Fe(BR06)}$. As shown in Figure $3d$, the results for a BH with
527: $a_*=0.95$ ($R_{\rm ISCO} = 1.973R_{\rm g}$) are very similar. Again,
528: for $H/R = 0.1$ the disk thickness $H$ decreases gradually near and
529: within the ISCO.
530:
531: We hasten to add that this is a very simple model of an accretion
532: disk, perhaps too simple to address ``surface phenomena'' such as
533: fluorescent iron line emission. Modulo this important caveat it seems
534: that, for reasonable values of the model parameters, the iron line
535: emission does not necessarily end at the ISCO, nor does it vary with
536: radius outside the ISCO with anything like the functional form assumed
537: in current fits of iron line data (e.g., BR06).
538:
539: \subsection{Viscous Stress Near the ISCO}
540:
541: Figure $4$ shows the vertically integrated stress $2 H \alpha P$ for a
542: non-spinning BH. As shown in Figure $4a$, all the models
543: corresponding to a very thin disk are in close agreement with the
544: standard model, i.e., the stress nearly vanishes at the ISCO even
545: though we do not require this of the model. For the thicker disk
546: shown in Figure $4b$, the stress near and inside the ISCO increases,
547: the effect becoming more important for larger values of \al.
548: Interestingly, for $\alpha=0.01$, the magnitude of the peak
549: stress is actually smaller than that predicted by the standard disk
550: model.
551:
552: As shown in Figure $5$, our models for a spinning black hole display
553: essentially this same dependence of stress on $H/R$ and \al. In both
554: Figures 4 and 5, the presence of a non-zero viscous stress inside the
555: ISCO implies a contribution to the observed spectrum that is not
556: accounted for in the standard disk model. In \S4 we investigate the
557: magnitude of this effect.
558:
559: We now consider the effect of \al\ and disk thickness on the
560: eigenvalue $j$ (\S2.2), which is the specific angular momentum
561: delivered to the black hole by the infalling matter. In the standard
562: disk model, $j$ is the Keplerian specific angular momentum at the ISCO
563: because (i) matter is assumed to orbit at the Keplerian velocity and
564: (ii) the stress is assumed to vanish inside the ISCO. Neither
565: assumption is made in our hydrodynamic models, and it is therefore of
566: interest to consider how much the calculated values of $j$ differ from
567: the standard value. Table~1 summarizes the values of $j$ for our
568: different models. We note that $j$ decreases with increasing $\HR$ and
569: \al. That is, as the disk gets thicker or as \al\ increases, more
570: angular momentum is removed before matter falls into the BH. However,
571: the effects are quite small, and the deviations are less than 1\% in
572: all cases.
573:
574: \subsection{Dissipation Inside The ISCO}
575:
576: In the previous section we showed that the stress at the ISCO is small,
577: but non-zero, and that it increases with disk thickness and \al. We now
578: consider the energy dissipation profiles of our model disks for
579: different values of \al\ and $\HR$. Figure $6$ shows the quantity,
580: \begin{equation}
581: R \frac{dL}{dR}=\frac{dL}{d \ln (R)}=\frac{-\dot{M}\alpha c_{\rm
582: s}^2 R^2}{v_{\rm R}} \frac{d\Omega}{dR}= 4 \pi R^2 D(R) \,,
583: \end{equation}
584: as a function of $R$. Here $L$ is the luminosity and $D(R)$ the energy
585: dissipated per unit time per unit surface area of the disk. Figures
586: $6a$ and $6b$ show $R dL/dR$ vs $R$ for $a_*=0$, while Figures $6c$
587: and $6d$ show the results for $a_*=0.95$. The solid black lines show
588: the standard disk model with zero torque at the ISCO. For the thin
589: disk with $\HR =0.01$ and for all values of \al, our models are
590: indistinguishable from the standard model, which thus provides an
591: excellent description of the flow in this case. However, for the
592: thicker disk with $\HR=0.1$, our numerical models deviate somewhat
593: from the standard disk model. We note in particular that larger
594: values of \al\ are associated with more dissipation near the ISCO and
595: larger deviations from the standard disk model.
596:
597: In Table 1 we summarize the total luminosities of the different models
598: for a given mass accretion rate \mdot. We note that none of the
599: luminosities of our models deviates by more than 4\% from that of the
600: standard model.
601:
602:
603: \begin{center}
604: \begin{tabular}{ccccccc}
605: \multicolumn{7}{c}{Table 1} \\ & \\
606: \multicolumn{7}{c}{Luminosities and Angular Momentum Eigenvalues
607: of the Numerical Disk Models} \\ \\
608: \hline
609: \hline
610: \multicolumn{1}{c}{$a_*$}&\multicolumn{1}{c}{H/R}&\multicolumn{1}{c}
611: {$\alpha$}&\multicolumn{1}{c}{$L_{\rm TOTAL}$}
612: &\multicolumn{1}{c}{$L_{\rm TOTAL(STD)}$}&\multicolumn{1}{c}{$
613: j$}&\multicolumn{1}{c}{$j_{\rm STD}$}\\
614: \multicolumn{1}{c}{}&\multicolumn{1}{c}{}&\multicolumn{1}{c}{}&\multicolumn{1}{c}{${\rm
615: (\dot{M}c^2)}$}
616: &\multicolumn{1}{c}{${\rm (\dot{M} c^2)}$}&\multicolumn{1}{c}{}
617: &\multicolumn{1}{c}{} \\
618: \hline
619: 0&0.01 & 0.01 & 0.0624 & 0.0625 & 3.6744 & 3.6742 \\
620: & & 0.1 & 0.0625 & & 3.6735 & \\
621: & & 0.2 & 0.0626 & & 3.6727 & \\
622: & & variable & 0.0626 & & 3.6730 & \\
623: \hline
624: &0.1 &0.01 & 0.0610 & & 3.6839 & \\
625: & &0.1 & 0.0633 & & 3.6609 & \\
626: & &0.2 & 0.0650 & & 3.6456 & \\
627: & &variable&0.0646 & & 3.6518 & \\
628: \hline
629: 0.95 & 0.1&0.01 & 0.2091 & 0.2144 & 2.3372 & 2.3311 \\
630: & &0.1 & 0.2171 & & 2.3237 & \\
631: & &0.2 & 0.2227 & & 2.3146 & \\
632:
633: \hline
634: \multicolumn{7}{l}{The subscript STD refers to the standard thin
635: disk model.} \\
636:
637: \end{tabular}
638: \end{center}
639:
640:
641:
642:
643:
644: \section{DISK SPECTRA AND THE EFFECT ON BH SPIN ESTIMATION}
645:
646: In the standard disk model, the viscous dissipation is assumed to vanish
647: at the ISCO. As a result, the emitted flux also vanishes at the ISCO,
648: and no radiation is emitted from the region of the flow between the ISCO
649: and the event horizon. For a given BH mass, the radius of the ISCO is a
650: well-known and monotonically decreasing function of $a_*$, e.g., for
651: $a_*$= 0, 1, the ISCO is located at $ 6 R_{\rm g}$, $1 R_{\rm g}$,
652: respectively. As discussed in Zhang et al. (1997), S06 and M06, the
653: radius $R_{\rm in}$ of the inner edge of the disk can be estimated from
654: observations. For a BH of known mass, this radius can be expressed in
655: units of $R_{\rm g}$, and if the disk inner edge is located at the ISCO,
656: then $R_{\rm in}/R_g$ determines the spin parameter $a_*$.
657:
658: From the calculations presented in this paper, we see that for a
659: very thin disk ($\HR=0.01$) the viscous dissipation does indeed
660: become negligible inside the ISCO and the dissipation profile $R
661: dL/dR$ is identical to that predicted by the standard disk model.
662: Thus for such systems we expect our estimates of BH spin to be quite
663: accurate. However, we do notice a difference for thicker disks with
664: say $H/R \sim0.1$. Using the standard disk model to fit the observed
665: spectra of these systems will lead to an error in our estimate of
666: the radius of the ISCO. We now try to quantify this error.
667:
668: For each of our disk solutions, we have calculated the emitted
669: spectrum assuming that the disk emits like a blackbody at each
670: radius. The temperature profile $T(R)$ of the disk surface can be
671: calculated from $dL/dR$ using:
672: \begin{equation}
673: (1-f)\frac{dL}{dR}= 4\pi \sigma R T^4(R) \; ,
674: \end{equation}
675: where $\sigma$ is the Stefan-Boltzmann constant. This can be used to
676: calculate the observed spectrum of the disk by integrating over the
677: entire disk:
678: \begin{equation}
679: F_{\rm \nu}= \frac{2 \pi \cos i}{D^2}\int_{R_{\rm inner}}^{R_{\rm
680: out}} \frac{2 h
681: \nu^3 R dR}{c^2 e^{\left( h\nu/ k T(R)) -1\right )}} \; ,
682: \end{equation}
683: where $h$ is the Planck constant, $c$ the speed of light, $k$ the
684: Boltzman constant, $D$ the distance, $i$ the angle of inclination,
685: $\nu$ the frequency, and $R_{\rm {inner}}$ the radius of the inner
686: boundary of the disk, near the event horizon.
687:
688: Figure $7$ shows our calculated spectra for a BH with mass $M=10
689: M_{\odot}$ and distance $D=10$ kpc. In each panel the solid curve
690: shows the spectrum from a standard disk model with the appropriate
691: pseudo-Newtonian potential. As before, we have considered three
692: constant values of $\alpha$: 0.01, 0.1 and 0.2, for both the spinning
693: and non-spinning cases, and an additional variable-\al\ model for the
694: non-spinning case.
695:
696: Figures $7a$ and $7b$ show the calculated spectra for the case of a
697: non-spinning BH. For $H/R=0.01$, we see that the calculated spectra
698: for all four models of \al\ overlap with the spectrum calculated via
699: the standard disk model. Therefore, we can conclude that, for such
700: very thin disks, the standard disk model is a very good approximation
701: and that the choice of \al\ cannot be a major source of error in
702: estimating BH spin. The different curves are more distinct in the case
703: of a thicker disk with $\HR=0.1$ (magenta, blue, red and green lines
704: show the \al~= 0.01, 0.1, 0.2, and variable-\al\ models). The
705: differences are especially noticeable at high photon energies, where
706: larger values of \al\ give higher fluxes. Figures $7c$ and $7d$ show
707: $H/R=0.1$ disk spectra for $a_*=0.8$ and 0.95.
708:
709: To estimate how the spectral distortions might affect BH spin
710: determination, we produced spectral data files for our models using
711: an $RXTE$ response file and analyzed the data with XSPEC {\it
712: version 12.2.0}. These ``fake'' data files were fitted with the
713: XSPEC model $Diskpn$ (Gierlinski et al. 1999), which uses the
714: standard disk model with the PW80 potential and a zero torque
715: boundary condition at the ISCO. $Diskpn$ has three fit parameters:
716: \Tmax, $R_{\rm in}/R_{\rm g}$ and normalization $K=M^2\cos
717: i/D^2\beta$, where $M$ is the mass, $D$ the distance, $i$ the angle
718: of inclination, and $\beta$ the color correction factor. We are
719: interested in the case when the inner edge of the disk coincides
720: with the ISCO. Thus, since $Diskpn$ considers a non-spinning BH, we
721: set $R_{\rm in}= 6 R_{\rm g}$. The constant $K$ can then be
722: rewritten as
723: \begin{equation}
724: K=\left( \frac{R_{\rm in}}{8.86 \times 10^6 \, {\rm cm}} \right )^2
725: \left ( \frac{D}{10 \, {\rm kpc}} \right )^{-2} \frac{\cos i}{\beta} \; .
726: \end{equation}
727: In the above expression, $8.86 \times 10^6$ cm corresponds to $6 R_{\rm
728: g}= R_{\rm ISCO}$ for a non-spinning black hole with $M= 10
729: M_\odot$. From the value of $K$ obtained from spectral fitting, one can
730: calculate $R_{\rm in}$ for each model using equation (27). Using this
731: value of $R_{\rm in}$, one can then calculate the BH spin for which the
732: ISCO would be located at that radius. This is the spin that one infers
733: from the fake spectral data, under the assumption that the standard disk
734: model is correct. Since the model was calculated with full viscous
735: hydrodynamics as described in earlier sections, the spin value derived
736: assuming the standard disk model will be different from the true BH spin
737: ($a_*=0$ in this case). The difference between the two values
738: represents the error in the spin estimate, $\Delta a_*$, caused by our
739: use of the simplified standard disk model.
740:
741: The results of this analysis are summarized in Table 2. The results
742: correspond to $M= 10 M_\odot$, $D$= 10 kpc, $\cos i=1$ and
743: $\beta=1$.
744:
745:
746:
747: \begin{center}
748: \begin{tabular}{cccc}
749: \multicolumn{4}{c}{Table 2} \\ & \\
750: \multicolumn{4}{c}{Errors in Spin Estimation of a Non-spinning BH} \\ \\
751: \hline
752: \hline
753: \multicolumn{1}{c}{$H/R$}&\multicolumn{1}{c}{$\alpha$}&\multicolumn{1}{c}{$R_{\rm
754: in}/(8.86 \times 10^6 \, {\rm cm})$}&\multicolumn{1}{c}{$\Delta a_*$} \\
755: \hline
756: 0.01 & 0.01 &1.011 &-0.011 \\
757: & 0.1 & 1.008 & -0.008 \\
758: & 0.2 & 1.006 & -0.006 \\
759: & variable& 1.007 & -0.004 \\
760: \hline
761: 0.1 &0.01 & 1.021 & -0.019 \\
762: &0.1 & 0.960 & 0.037 \\
763: &0.2 & 0.920 & 0.074 \\
764: &variable&0.879 & 0.060 \\
765:
766: \hline
767: \end{tabular}
768: \end{center}
769:
770:
771: Figure $8a$ shows the results in more detail for $\HR=0.01$, 0.02, 0.04,
772: 0.06, 0.08 and 0.1. Figure $8b$ shows the results for the variable-\al\
773: model as a function of disk thickness. We see that the error is larger
774: for thicker disks and also for larger values of \al. However, even for
775: the thickest case we considered, $\HR=0.1$, and the largest value of
776: $\alpha=0.2$, the BH spin is overestimated by less than $0.1$. Thus, in
777: the case of a non-spinning BH the error is quite modest when one
778: considers, for example, that both the radius of the ISCO and the binding
779: energy at the ISCO differ only slightly (by 6\%) for a BH with $a_* =
780: 0.1$.
781: %We also notice that for $\HR=0.01$ the spin has a
782: %small non-zero negative value.
783: %This effect
784: %is not visible in the figures, where all the different \al\ models
785: %seem to coincide with the standard disk model.
786:
787: Though a similar standard pseudo-Kerr $XSPEC$ model is not available
788: for fitting our spinning BH model spectra, it is still possible to
789: estimate $\Delta a_*$ by calculating the model luminosities.
790: Gierlinski et al. (1999) showed that, for a non-spinning BH with the
791: PW80 potential, one can write:
792: \begin{equation}
793: L=\frac{1}{16} \dot{M}c^2 =35.7 \frac{\pi
794: \sigma}{\beta^4}R_{\rm in}^2 T_{\rm max}^4 \; ,
795: \end{equation}
796: where $L$ is the luminosity, $\sigma$ the Stefan-Boltzmann constant,
797: $\beta$ the color correction factor, $R_{\rm in}$ the radius of the
798: inner edge of the disk, and $T_{\rm max}$ the peak temperature of
799: the disk. Therefore, instead of calculating the multicolor blackbody
800: spectrum of our models and fitting them with XSPEC, we could simply
801: compare each hydrodynamic model with the corresponding standard disk
802: model with the same $T_{\rm max}$. This gives the following
803: estimate for the effective disk inner radius $R_{\rm in}$ of any
804: given hydrodynamic model,
805: \begin{equation}
806: R_{\rm in}^2= \frac{ L_{\rm model}}{L_{\rm standard \; disk}}
807: R_{\rm ISCO}^2 \; ,
808: \end{equation}
809: where $L_{\rm model}$ is the luminosity of the model, $L_{\rm
810: standard \; disk}$ is the luminosity of the standard disk with the
811: same value of $T_{\rm max}$, and $R_{\rm ISCO}$ is the radius of the
812: ISCO. The value of $R_{\rm in}$ obtained using equation (29) may
813: then be used to calculate $\Delta a_*$, as before.
814:
815:
816: Figure 9a shows $\Delta a_*$ values calculated using both the full
817: spectral fitting method via equation (27) and the simpler
818: luminosity-temperature method described by equation (29). We see that
819: the results are very close, indicating that the second method is a
820: good proxy for the more detailed spectral method.
821:
822: For a spinning BH, equation (28) can be generalized to
823: \begin{equation}
824: L=\epsilon \dot{M}c^2 \sim c_0 \frac{\pi \sigma}{\beta^4}R_{\rm in}^2
825: T_{\rm max}^4 \; ,
826: \end{equation}
827: where $\epsilon$ is the spin-dependent efficiency of the BH, and
828: $c_0$ is a constant. Therefore, equation (29) can again be used to
829: estimate the effective $R_{\rm in}$ and this can be used to obtain
830: an estimate of the BH spin.
831:
832:
833: Figure $9b$ shows $\Delta a_*$ for spinning BHs using this method.
834: We show results for $a_*$ = 0.7, 0.8, 0.9 and 0.95, and $\HR=0.1$.
835: For a given disk thickness and \al, we see that the error in the
836: spin estimate becomes smaller as the spin of the BH increases. For
837: $a_*=0.95$, the maximum error is only $\sim 0.01$.
838:
839:
840: \section{DISCUSSION}
841:
842: In this paper we studied the properties of a simple hydrodynamic model
843: of an accretion disk using the $\alpha$ prescription for viscosity. We
844: considered models with finite thicknesses $H/R$ and different values
845: of $\alpha$. Our aim was to investigate how much the hydrodynamic
846: models of thin disks deviate from the idealized ``standard disk
847: model'' which assumes a vanishing torque at the innermost stable
848: circular orbit (ISCO).
849:
850: We find that the deviations of the viscous hydrodynamic models from
851: the standard disk model increase with increasing $H/R$ and increasing
852: $\alpha$. However, even for $H/R=0.1$ and $\alpha~= 0.2$, the largest
853: values we tried for our thin disk calculations, the deviations remain
854: modest. This is illustrated in Figures 4 and 5, which show how the
855: stress profile deviates from that of the idealized standard disk
856: model, and also in Figure 6, which compares the profiles of the
857: viscous energy dissipation rate $R dL/dR$, Figure 7, which shows the
858: multicolor blackbody spectra of the models, and Table 1, which gives
859: some quantitative results. In all cases, we see that the detailed
860: hydrodynamic models match the standard disk model quite closely.
861:
862: We were motivated to do this study because we and others have used the
863: standard disk model to fit the continuum spectra of BH X-ray binaries
864: in the thermal state in order to estimate the spins of the BHs. How
865: much error do we expect in the estimated spin values as a result of
866: the fact that a real disk deviates from the standard disk model? At
867: least for the simple hydrodynamic models we have considered in this
868: paper, the answer is that the errors are quite modest.
869:
870: Quantitative results are given in Table 2 and Figures 8 and 9. The
871: error $\Delta a_*$ in the derived estimate of BH spin is at most
872: $\sim0.1$ in the case of a non-spinning BH and is much less for
873: rapidly spinning BHs. These errors are comparable to or smaller
874: than the errors that arise from uncertainties in our estimates of
875: mass, distance and disk inclination (S06, M06).
876:
877: While these results are very encouraging for our program to estimate
878: BH spin through fitting the continuum spectra of BH accretion disks in
879: the thermal state, we must note some caveats. First and foremost, we
880: have considered a highly simplified toy hydrodynamic model with an
881: $\alpha$ prescription for viscosity. Real disks doubtless have
882: magnetic fields, and the stresses associated with these fields
883: probably do not behave like microscopic viscosity. Indeed, it is
884: precisely this argument that has been used by Krolik (1999),
885: Gammie (1999) and HK02 to question the zero-torque boundary condition
886: at the ISCO. On the other hand, Paczy\'nski (2000) makes an equally
887: persuasive argument (based on the angular momentum conservation
888: equation) that, so long as the shear stress is smaller than the
889: pressure, a thin disk will always satisfy the zero-torque condition.
890:
891: In an attempt to include some of the effects of magnetic fields, we
892: have considered a model in which we allowed $\alpha$ to vary with
893: radius (see eq. 22) in such a manner as to closely mimic the effective
894: $\alpha$ obtained by HK02 from their MHD simulations. Even though in
895: this model $\alpha$ increases rapidly with decreasing radius,
896: especially inside the ISCO, we found that none of our results changed.
897: Based on this finding we cautiously suggest that the inclusion of
898: magnetic fields may not significantly alter our conclusions.
899:
900: One question that needs to be addressed is why our results differ so
901: much from those obtained by HK02. From MHD simulations of magnetized
902: gas accreting in a PW80 potential, those authors concluded that the
903: vertically integrated magnetic stress increases monotonically with
904: decreasing radius all the way through and inside the ISCO. This is
905: dramatically different from the behavior we find, as a comparison of
906: HK02's Fig. 10 with our Fig. 4 shows. A likely explanation is that we
907: have limited our study to {\it thin} disks ($H/R=0.01, ~0.1$) in which
908: we simulated strong cooling by choosing a small value for the
909: advection parameter $f$ (see the discussion below eq. 19). HK02, by
910: contrast, had no cooling in their MHD simulation, so their gas
911: retained whatever energy was generated through shocks, making their
912: disk thicker.
913:
914: In order to verify that this difference is important, we calculated
915: models with larger values of $f$ using our viscous hydrodynamic code.
916: It is hard to know what effective value of $f$ is most appropriate to
917: match the HK02 simulation. Nominally, theirs was a fully
918: advection-dominated accretion flow, since they had no cooling at all;
919: this means that their simulation corresponded to $f=1$. However, we
920: do not know how well their code conserved energy. Therefore, we
921: calculated three models with $f=1$, 0.5 and 0.25, all with the
922: variable $\alpha$ prescription (eq. 22) which most closely matches
923: their stress profile. Figure 10 shows the resulting stress profiles.
924: We see that these advection-dominated models do exhibit a
925: monotonically increasing stress inward, exactly as found by HK02
926: (their Fig. 10). The stress profiles are very different from those we
927: find for cooling-dominated thin disks (our Figs. 4 and 5). Thus, we
928: tentatively suggest that a large part of the difference between the
929: results we find in this paper and those obtained by HK02 is related to
930: the differing treatments of the energy equation of the gas, viz.,
931: cooling-dominated thin disk regime versus advection-dominated thick
932: disk regime. In other words, we confirm the original insight of
933: Abramowicz \& Kato (1989), Paczy\'nski (2000) and Afshordi \&
934: Paczy\'nski (2003) on the strong relation between disk thickness and
935: the stress at the ISCO. However, only a detailed MHD study of an
936: accretion disk with significant cooling can tell for sure if this
937: interpretation is correct, and to our knowledge nobody has carried out
938: such a study.
939:
940: Another limitation in our work is that we used a Newtonian model and
941: we simplified the thermodynamics of the gas in the disk via the
942: advection parameter $f$ (see eq. 10). However, doing the calculations
943: in general relativity with full radiation thermodynamics will, we
944: believe, introduce modifications only of order unity. The changes
945: will be larger for a spinning BH, which we modeled with the
946: Mukhopadhyay (2002) model, compared to a non-spinning hole (PW80
947: potential), but we think the error will still be only of order unity.
948: Therefore, calculating these effects in more detail will not greatly
949: alter our qualitative conclusion that the standard disk model is
950: adequate so long as the disk geometrically thin. Nevertheless,
951: it would be useful to extend this work using a more complete set of
952: disk equations, such as those employed in the study of slim disks
953: (Abramowicz et al. 1988), and with the inclusion of general
954: relativity (e.g., Abramowicz, Lanza \& Percival 1997).
955:
956: Note that we employed the two pseudo-Newtonian potentials mentioned
957: above in the work reported here merely to obtain a ballpark estimate
958: of the error associated with the zero-torque approximation. When we
959: actually fit data to estimate the spin parameters of BHs (e.g., S06,
960: M06), we use a detailed model (Li et al. 2005) which assumes the Kerr
961: metric and includes all special relativistic and general relativistic
962: effects.
963:
964: In our work on BH spin (S06, M06), we limited ourselves to disks with
965: luminosities less than 30\% of Eddington, which corresponds to
966: vertical thicknesses $H/R<0.1$. The present study shows that this was
967: a reasonable choice. For $H/R \leq 0.1$, the effects of gas physics
968: and finite vertical thickness in our hydrodynamic models are not
969: serious. Equally clearly, for thicker disks with $H/R$ much greater
970: than 0.1, the effects will be large; e.g., see Figure 10. Therefore,
971: one should be cautious about applying the standard disk model to disks
972: more luminous than 30\% of Eddington. For this reason, we
973: believe the results obtained by Middleton et al. (2006) for the spin
974: of the microquasar GRS 1915+105 should be taken with caution.
975:
976: Strong observational evidence that fitting the X-ray continuum is a
977: promising way to estimate black hole spin comes from a long history of
978: fitting the broadband spectra of black hole transients using the simple
979: non-relativistic multicolor disk model (Mitsuda et al. 1984; Makishima
980: et al. 1986), which returns the temperature $T_{\rm in}$ at the
981: inner-disk radius $R_{\rm in}$. In their classic review, Tanaka \&
982: Lewin (1995) give examples of the steady decay (by factors of 10--100)
983: of the thermal flux of transient sources during which $R_{\rm in}$
984: remains constant. They remark that the constancy of $R_{\rm in}$
985: suggests that it is related to the radius of the ISCO. More recently,
986: this evidence for a constant inner radius in the thermal state has been
987: presented for a number of sources via plots showing that the bolometric
988: luminosity of the thermal component is approximately proportional to
989: $T_{\rm in}^4$ (Kubota \& Makishima 2001; Kubota \& Makishima 2004; Abe
990: et al. 2005; McClintock et al. 2007). In short, these non-relativistic
991: analyses, which ignore spectral hardening (Davis et al. 2006), provide
992: evidence for the presence of a stable radius, although they obviously
993: cannot provide a secure value for the radius of the ISCO or even
994: establish that the stable radius is the ISCO.
995:
996: We now consider the iron-line method of estimating spin. In this
997: method, it is assumed that the line emission ceases abruptly at the
998: ISCO, so an important question is whether or not the gas inside the
999: ISCO will fluoresce (Reynolds \& Begelman 1997). The possibility of
1000: line emission from inside the ISCO is usually discounted on the
1001: grounds that the density will fall suddenly inside the ISCO, thus
1002: causing a sudden increase in the ionization parameter (Fabian 2007,
1003: and references therein). Alternatively, and to the same effect, it is
1004: argued that emissivity is related to the ``gravity parameter''
1005: (Nayakshin 2000, 2002), and should depend on $H$. We see in Figure
1006: $3a$ that the density dependent function $\rho(R) R^3$ does become
1007: negligible inside the ISCO for the non-spinning BH. However, that is
1008: not the case for a fast spinning BH with $a_*=0.95$ (Figure $3b$) even
1009: for a small disk thickness of $\HR=0.1$. Also, the radial dependence
1010: of $H$ shown in Figures $3c$ and $3d$ implies that there should be
1011: emission from the inner region unless the disk is very thin
1012: ($\HR=0.01$).
1013:
1014:
1015: An additional complication for iron line modeling is that the
1016: emissivity is assumed to vary as a broken power-law, with the maximum
1017: emission occurring exactly at the ISCO (e.g., BR06). Looking at
1018: Figure 3, such an ad hoc model would be hard to justify if the
1019: emissivity has anything to do with gas density or disk thickness. In
1020: contrast, the continuum-fitting model has the merit that it makes use
1021: of a physically motivated profile of disk emission $R dL/dR$ which can
1022: be calculated from first principles in the standard disk model and
1023: which continues to be valid even in the more general hydrodynamic
1024: models described in this paper (Figs. 6, 7).
1025: %It would thus appear
1026: %that, at the present time, the continuum fitting method is the more
1027: %robust approach to estimating BH spin. Apart from being grounded better
1028: %in physics, the method seems reliable up to $H/R \sim 0.1$, whereas the
1029: %iron line method probably becomes unreliable for much thinner disks.
1030:
1031: This paper has focused on only one aspect of BH spin estimation, viz.,
1032: the validity of assumptions made in various methods of spin
1033: determination regarding the hydrodynamical properties of the accretion
1034: disk. Of course, a successful determination of spin needs more than a
1035: valid disk model. It also requires high quality data and accurate
1036: determination of secondary system parameters. A discussion of these
1037: issues is beyond the scope of this paper, and the reader is referred
1038: to appropriate papers in the literature (e.g., M06; BR06).
1039:
1040: \acknowledgements The authors thank Niayesh Afshordi and Jonathan
1041: McKinney for discussions and useful suggestions. We dedicate this
1042: paper to Bohdan Paczy\'nski for his amazing insights in accretion
1043: theory and in numerous other areas of astrophysics.
1044:
1045: \newpage
1046:
1047: \begin{thebibliography}{}
1048:
1049: \bibitem[Abe et al.(2005)]{A05} Abe, Y., Fukazawa, Y., Kubota, A.,
1050: Kasama, D., \& Makishima, K. 2005, PASJ, 57, 629
1051:
1052: \bibitem[Abramowicz et al.(1988)]{Abr88} Abramowicz, M. A., Czerny,
1053: B., Lasota, J. P., \& Szuszkiewicz, E. 1988, ApJ, 332, 646
1054:
1055: \bibitem[Abramowicz \& Kato(1989)]{Abr89} Abramowicz, M. A., \& Kato,
1056: S. 1989, ApJ, 336, 304
1057:
1058: \bibitem[Abramowicz et al.(1997)]{Abr97} Abramowicz, M. A., Lanza,
1059: A., \& Percival, M. J. 1997, ApJ, 479, 179
1060:
1061: \bibitem[Afshordi \& Paczynski(2003)]{Afs03} Afshordi, N., \&
1062: Paczy\'nski, B. 2003, ApJ, 592, 354
1063:
1064: \bibitem[Balbus \& Hawley (1991)]{BH91} Balbus, S. A., \& Hawley,
1065: J. F. 1991, ApJ, 376, 214
1066:
1067: \bibitem[Brenneman \& Reynolds (2006)]{BR97} Brenneman, L. W., \&
1068: Reynolds, C. S. 2006, ApJ, 652, 1028 (BR06)
1069:
1070: \bibitem[Chen et al.(1997)]{Che97} Chen, X., Abramowicz, M. A.,
1071: \& Lasota, J. P. 1997, ApJ, 476, 61
1072:
1073: \bibitem[Chen \& Taam(1993)]{Che93} Chen, X., \& Taam, R. E. 1993,
1074: ApJ, 412, 254
1075:
1076: \bibitem[Davis et al. (2005)]{D05} Davis, S. W., Blaes, O. M., Hubeny,
1077: I., Tuner, N. J. 2005, ApJ, 621, 372
1078:
1079: \bibitem[Davis et al.(2006)]{D06} Davis, S. W., Done, C., \& Blaes,
1080: O. M. 2006, ApJ, 647, 525
1081:
1082: \bibitem[Frank et al.(2002)]{F02} Frank, J., King, A., Raine,
1083: D. J. 2002, Accretion Power in Astrophysics (Cambridge: Cambridge
1084: Univ. Press)
1085:
1086: \bibitem[Fabian (2007)]{F07} Fabian, A. C. 2007, in Black Holes: from
1087: Stars to Galaxies, Proc. IAU Symp. 238, eds V. Karas, G. Matt, in
1088: press (astro-ph/0612435)
1089:
1090: \bibitem[Gammie 1999]{G99} Gammie, C. F. 1999, ApJ, 522, 57
1091:
1092: \bibitem[Gierlinski et al. (1999)]{Gi99} Gierlinski, M., Zdziarski,
1093: A. A, Poutanen, J., Coppi, P. S., Ebisawa, K., \& Johnson, W. N.
1094: 1999, MNRAS, 309, 496
1095:
1096: \bibitem[Hawley \& Krolik (2002)]{HK02} Hawley, J. F., \& Krolik,
1097: J. H. 2002, ApJ, 566, 164 (HK02)
1098:
1099: \bibitem[Kato et al.(1988)]{Kat88} Kato, S., Honma, F., \&
1100: Matsumoto, R. 1988, MNRAS, 231, 37
1101:
1102: \bibitem[Krolik 1999]{K99} Krolik, J. H. 1999, ApJL, 515, 73
1103:
1104: \bibitem[Krolik \& Hawley(2002)]{KH02} Krolik, J. H., \& Hawley, J. F.
1105: 2002, ApJ, 573, 754
1106:
1107: \bibitem[Kubota(2004)]{KM04} Kubota, A., \& Makishima, K. 2004, ApJ,
1108: 601, 428
1109:
1110: \bibitem[Kubota(2001)]{KM01} Kubota, A., Makishima, K., \& Ebisawa,
1111: K. 2001, ApJ, 560, L147
1112:
1113: \bibitem[Li et al.(2005)]{Li05} Li, L.-X., Zimmerman, E. R., Narayan,
1114: R., \& McClintock, J. E. 2005, ApJS, 157, 335
1115:
1116: \bibitem[Makishima et al.(1986)]{M86}
1117: Makishima, K., Maejima, Y., Mitsuda, K., Bradt, H. V., Remillard, R. A.,
1118: Tuohy, I. R., Hoshi, R., \& Nakagawa, M. 1986, ApJ, 308, 635
1119:
1120: \bibitem[McClintock et al.(2007)]{M07} McClintock, J. E., Remillard,
1121: R. A., Rupen, M. P., Torres, M. A. P., Steeghs, D., Levine, A. M., \&
1122: Orosz, J. A. 2007, ApJ, submitted (astro-ph/0705.1034)
1123:
1124: \bibitem[McClintock et al.(2006)]{M06} McClintock, J., E., Shafee, R.,
1125: Narayan, R., Remillard, R. A., Davis, S. W., Li, L.-X. 2006, ApJ,
1126: 652, 518 (M06)
1127:
1128: \bibitem[Middleton et al.(2006)]{Mid06} Middleton, M., Done, C.,
1129: Gierli\'nski, M., \& Davis, S. W. 2006, MNRAS, 373, 1004
1130:
1131: \bibitem[Mitsuda et al.(1984)]{M84} Mitsuda, K. et al. 1984, PASJ, 36,
1132: 741
1133:
1134: \bibitem[Muchotrzeb \& Paczy\'nski(1982)]{Pac82} Muchotrzeb, B. \&
1135: Paczy\'nski, B. 1982, Acta Astron., 32, 1
1136:
1137: \bibitem[Mukhopadhyay 2002]{M02}Mukhopadhyay, B. 2002, ApJ, 581, 427
1138:
1139: \bibitem[Narayan et al.(1997)]{Nar97} Narayan, R., Kato, S., \&
1140: Honma, F. 1997, ApJ, 476, 49
1141:
1142: \bibitem[Narayan \& Popham(1993)]{Nar93} Narayan, R., \& Popham,
1143: R. 1993, Nature, 362, 820
1144:
1145: \bibitem[Narayan \& Yi (1994)]{NY94} Narayan, R., Yi, I., ApJL, 428,
1146: 13
1147:
1148: \bibitem[Nayakshin et al. (2002)]{N00}Nayakshin, S., Kazanas, D.,
1149: Kallman, T. 2000, ApJ, 537, 833
1150:
1151: \bibitem[Paczynski(2000)]{Pac00} Paczy\'nski, B. 2000, astro-ph/0004129
1152:
1153: \bibitem[Paczy\'nski \& Bisnovatyi-Kogan(1981)]{Pac81} Paczy\'nski, B.,
1154: \& Bisnovatyi-Kogan, G. 1981, Acta Astron., 31, 283
1155:
1156: \bibitem[Paczynski \& Wiita (1980)]{PPW80} Paczy\'nski, B., \& Wiita,
1157: P. J. 1980, A\&A, 343, 325 (PW80)
1158:
1159: \bibitem[Popham \& Narayan(1991)]{Pop91} Popham, R., \& Narayan,
1160: R. 1991, ApJ, 370, 604
1161:
1162: \bibitem[Quataert \& Narayan (2000)]{QN00} Quataert, E., \& Narayan,
1163: R. 1999, ApJ, 516, 399
1164:
1165: \bibitem[Remillard \& McClintock (2006)]{RM06} Remillard, R. A.,
1166: McClintock, J. E. 2006, ARAA, 44, 49
1167:
1168: \bibitem[Reynolds \& Begelman (1997)]{RB97} Reynolds, C. S., \&
1169: Begelman, M. C. 1997, ApJ, 488, 109
1170:
1171: \bibitem[Ross \& Fabian (1993)] {RF93} Ross, R. R., Fabian, A. C., 1993, MNRAS, 261, 74
1172:
1173: \bibitem[Ross, Fabian \& Young (1999)]{RFY99}
1174: Ross, R. R., Fabian, A. C., Young, A. J., 1999, MNRAS, 306, 461
1175:
1176:
1177: \bibitem[Shafee et al.(2006)]{S06} Shafee, R., McClintock, J. E.,
1178: Narayan, R., Davis, S. W., Li, L.-X., \& Remillard, R. A. 2006,
1179: ApJL, 636, 113 (S06)
1180:
1181: \bibitem[Shakura \& Sunyaev (1973)]{SS73} Shakura, N. I., \& Sunyaev,
1182: R. A. 1973, A\&A, 24, 337
1183:
1184: \bibitem[Tanaka \& Lewin(1995)]{TL95} Tanaka, Y., \& Lewin,
1185: W. H. G. 1995, in X-ray Binaries, eds. W. Lewin, J. van Paradijs, \&
1186: E. van den Heuvel (Cambridge: Cambridge Univ. Press), p126
1187:
1188: \bibitem[Zhang, Cui \& Chen 1997]{ZCC} Zhang, S. N., Cui, W., \&
1189: Chen, W. 1997, ApJ, 482, L155
1190:
1191: \bibitem[\.{Z}ycki et al. (1994)]{Z94}
1192: \.{Z}ycki, P. T., Krolik, J. H., Zdziarski, A. A., Kallman, T. R.,
1193: 1994, ApJ, 437, 597
1194:
1195: \end{thebibliography}
1196:
1197: \newpage
1198: \begin{figure}
1199: \figurenum{1} \plotone{fig1.eps} \caption{ Disk parameters for a
1200: non-spinning BH: ($a$) sound speed, ($b$) radial velocity, ($c$)
1201: angular velocity, ($d$) density. In panels $a$, $b$, and $d$, the
1202: solid and dotted lines correspond to $\HR=0.01$ and $\HR=0.1$,
1203: respectively; where distinct, the magenta, blue, red and green lines
1204: represent $\alpha = 0.01$ , 0.1 and 0.2, and variable \al\ (eq. 22),
1205: respectively. In all three panels, the variable-\al~model is nearly
1206: coincident with the \al~= 0.1 model. In panel $c$, the Keplerian
1207: velocity is plotted as a solid red line. Because the angular-velocity
1208: profile of the four models are nearly identical, we represent them by
1209: a single dotted line. The radius of the ISCO, $R=6R_{\rm g}$, is
1210: indicated in all four panels by the vertical dashed line. All
1211: numerical values correspond to $G=c=M=1$, $\dot{M}=1$. In panel $c$,
1212: the unit of angular velocity is $(G M/c^3)^{-1}$. In panel $d$, the
1213: unit of $\rho$ is $c^6/(G^3 M^2)$.}
1214: \end{figure}
1215:
1216:
1217: \newpage
1218: \begin{figure}
1219: \figurenum{2} \plotone{fig2.eps} \caption{Similar to Figure 1, but for
1220: a spinning BH with $a_*=0.95$. The vertical dashed line shows the
1221: ISCO at R= $1.937 R_{\rm g}$. There is no variable-\al~model in
1222: this case (see text), and hence the green line is not present in
1223: these plots.}
1224: \end{figure}
1225:
1226:
1227: \newpage
1228: \begin{figure}
1229: \figurenum{3} \plotone{fig3.eps} \caption{Profiles of $\rho(R)R^3$
1230: (panels $a$ and $b$) and disk thickness $H$ (panels $c$ and $d$) in
1231: the inner regions. For the line types/colors defining the various
1232: models and the location of the ISCO, see Figures 1 and 2.
1233: Superimposed on our models in all of the panels is a thick black
1234: line that schematically represents the emissivity profile
1235: efficiency, $f_{\rm Fe}$, assumed in the iron line work in which the
1236: emissivity cuts off abruptly inside the ISCO and falls off as a
1237: steep power law outside the ISCO (eq. 23). Panel $a$ shows $\rho(R)
1238: R^3$ as a function of radius for a non-spinning BH. Panel $b$ shows
1239: a similar plot for $a_*=0.95$. Panels $c$ and $d$ show disk
1240: thickness $H$ as a function of radius for $a_*=0$ and 0.95, for
1241: disks with asymptotic values of $H/R = 0.01$ and 0.1. In panels $a$
1242: and $b$, the normalizations used for $H/R=0.01$ and $\HR=0.1$ are
1243: different.}
1244: \end{figure}
1245:
1246:
1247: \newpage
1248: \begin{figure}
1249: \figurenum{4} \plotone{fig4.eps} \caption{Vertically integrated
1250: stress $2 H \alpha P (\times 10^4)$ for a non-spinning BH. In
1251: both panels the standard disk model is plotted as a thick solid
1252: line. For the line types/colors defining the various models and
1253: the location of the ISCO, see Figure 1. ($a$) For $H/R = 0.01$,
1254: all four models are seen as indistinguishable from the standard
1255: model. ($b$) For the thicker disk, the models can be cleanly
1256: distinguished inside $R \sim 15R_{\rm g}$. All numerical values
1257: correspond to $G=c=M=1$, $\dot{M}=1$.}
1258: \end{figure}
1259:
1260:
1261: \newpage
1262: \begin{figure}
1263: \figurenum{5} \plotone{fig5.eps} \caption{Similar to Figure 4, but
1264: for $a_*$=0.95. As in Figure 2, there is no variable-\al~model, and
1265: the ISCO is located at $R=1.937 R_{\rm g}$.}
1266: \end{figure}
1267:
1268:
1269: \newpage
1270: \begin{figure}
1271: \figurenum{6} \plotone{fig6.eps} \caption{Rate of energy dissipation
1272: $R dL/dR$ as a function of radius $R$ for models of a non-spinning
1273: BH ($a$ and $b$) and a spinning BH ($c$ and $d$). For the line
1274: types/colors defining the various models and the location of the
1275: ISCO, see Figures 1 and 2. The models shown for the thinner disks
1276: coincide with the standard disk model, whereas the thicker disks
1277: deviate somewhat from the standard model.}
1278: \end{figure}
1279:
1280:
1281:
1282: \newpage
1283: \begin{figure}
1284: \figurenum{7} \plotone{fig7.eps} \caption{Spectra corresponding to
1285: the numerical disk models described in this paper for models of a
1286: non-spinning BH ($a$ and $b$) and spinning BHs with thicker disks
1287: for $a_* = 0.8$ ($c$) and $a_* = 0.95$ ($d$). For the line
1288: types/colors defining the various models, see Figures 1 and 2.
1289: Again, the different models are essentially indistinguishable in
1290: the case of the thin disk ($a$).}
1291: \end{figure}
1292:
1293:
1294: \newpage
1295: \begin{figure}
1296: \figurenum{8} \plotone{fig8.eps} \caption{($a$) Error in the spin
1297: estimate as a function of $\alpha$ for a BH with a true spin
1298: parameter of $a_{*} =0$. The quantity $\Delta a_*$ is equal to
1299: the value of $a_*$ obtained from fitting the model spectrum minus
1300: the true $a_{*}$. The different curves correspond to different
1301: relative thicknesses $H/R$ of the disk. ($b$) Error in the spin
1302: estimate for the variable-$\alpha$ profile as a function of disk
1303: thickness. In panel a, note the variable offset from zero error
1304: that occurs near \al~= 0.01, which is not visible in the previous
1305: figures for which the the different \al\ models nearly coincide
1306: with the standard disk model.}
1307: \end{figure}
1308:
1309:
1310: \newpage
1311: \begin{figure}
1312: \figurenum{9} \plotone{fig9.eps} \caption{($a$) Error in the spin
1313: estimate $\Delta a_*$ for a non-spinning BH, calculated using
1314: spectral fitting and eq.~(27) (solid lines) and from eq.~(29)
1315: (dotted lines). The agreement is very good, showing that the
1316: simpler approach via eq.~(29) is quite accurate. ($b$) $\Delta a_*$
1317: as a function of $\alpha$ for BHs with different values of $a_*$,
1318: calculated using equation~(29).}
1319: \end{figure}
1320:
1321: \newpage
1322: \begin{figure}
1323: \figurenum{10} \plotone{fig10.eps} \caption{The thick red solid line
1324: shows the vertically integrated stress profile as predicted by the
1325: standard disk model for a non-spinning BH (PW80 potential). The
1326: other lines show the stress profiles of three hydrodynamic disk
1327: models with advection parameter values $f=0.25$ (short-dashed line),
1328: $f=0.5$ (dotted line) and $f=1$ (long-dashed line). All three models
1329: use the variable $\alpha$ prescription (eq. 22), and their stress
1330: profiles have been scaled to match the standard model at
1331: $R=15R_g$. A logarithmic scale has been used to facilitate
1332: comparison with the MHD simulation result shown in Fig. 10 of
1333: HK02.}
1334: \end{figure}
1335:
1336: \end{document}
1337: