0806.2666/ms.tex
1: % mn2esample.tex
2: %
3: % v2.1 released 22nd May 2002 (G. Hutton)
4: %
5: % The mnsample.tex file has been amended to highlight
6: % the proper use of LaTeX2e code with the class file
7: % and using natbib cross-referencing. These changes
8: % do not reflect the original paper by A. V. Raveendran.
9: %
10: % Previous versions of this sample document were
11: % compatible with the LaTeX 2.09 style file mn.sty
12: % v1.2 released 5th September 1994 (M. Reed)
13: % v1.1 released 18th July 1994
14: % v1.0 released 28th January 1994
15: 
16: %\documentclass[referee,useAMS,usenatbib,usegraphicx]{mn2e}
17: \documentclass[useAMS,usenatbib,usegraphicx]{mn2e}
18: %\documentstyle[epsf, epsfig]{mn2e}
19: \usepackage{epsf, epsfig}
20: \usepackage{Times}
21: 
22: % If your system does not have the AMS fonts version 2.0 installed, then
23: % remove the useAMS option.
24: 
25: %
26: % useAMS allows you to obtain upright Greek characters.
27: % e.g. \umu, \upi etc.  See the section on "Upright Greek characters" in
28: % this guide for further information.
29: %
30: % If you are using AMS 2.0 fonts, bold math letters/symbols are available
31: % at a larger range of sizes for NFSS release 1 and 2 (using \boldmath or
32: % preferably \bmath).
33: %
34: % The usenatbib command allows the use of Patrick Daly's natbib.sty for
35: % cross-referencing.
36: %
37: % If you wish to typeset the paper in Times font (if you do not have the
38: % PostScript Type 1 Computer Modern fonts you will need to do this to get
39: % smoother fonts in a PDF file) then uncomment the next line
40: % \usepackage{Times}
41: 
42: %%%%% AUTHORS - PLACE YOUR OWN MACROS HERE %%%%%
43: 
44: \setlength\topmargin{-2.2pc}
45: \setlength\topmargin{0pt}
46: \def\be{\begin{equation}}
47: \def\ee{\end{equation}}
48: \def\mdot{\dot{m}}
49: \def\ergs{{\rm\,erg\,s^{-1}}}
50: \def\msun{M_{\odot}}
51: \def\ergscc{\rm \  \ erg \ cm^{-3} \ s^{-1}}
52: \def\gs{\rm \,g\,s^{-1}}
53: \def\ergscc{\rm \,erg\,cm^{-3}\,s^{-1}}
54: \def\ergs{\rm \,erg\,s^{-1}}
55: \def\be{\begin{equation}}
56: \def\ee{\end{equation}}
57: \catcode`\@=11 % This allows us to modify PLAIN macros.
58: \def\@versim#1#2{\vcenter{\offinterlineskip
59: \ialign{$\m@th#1\hfil##\hfil$\crcr#2\crcr\sim\crcr } }}
60: \def\lsim{\mathrel{\mathpalette\@versim<}}
61: \def\gsim{\mathrel{\mathpalette\@versim>}}
62: \def\mpy{M_\odot \ {\rm yr^{-1}}}
63: \newcommand{\arcdeg}{\ensuremath{^{\circ}}}
64: 
65: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
66: 
67: \title[X-rays from Quiescent Black Holes]{Origin of X-ray Emission from 
68: Transient Black Hole Candidates in Quiescence}
69: \author[Pszota et al.]
70: {Gabor Pszota$^{1}$\thanks{E-mail: pszotag@physics.purdue.edu}, 
71: Hui Zhang$^{2}$\thanks{E-mail: hzhang@shao.ac.cn}, 
72: Feng Yuan$^{2,3}$\thanks{E-mail: fyuan@shao.ac.cn},
73: Wei Cui$^{1}$\thanks{E-mail: cui@physics.purdue.edu}
74: \\
75: $^{1}$\/Department of Physics, Purdue University, West Lafayette, IN 47907\\
76: $^{2}$\/Shanghai Astronomical Observatory, Chinese Academy of Sciences, 
77: 200030 Shanghai \\
78: $^{3}$\/Joint Institute for Galaxy and Cosmology of Shanghai Astronomical 
79: Observatory and University of Science and Technology of China}
80: 
81: \voffset=-0.8in
82: 
83: \begin{document}
84: 
85: \date{Received 2008}
86: 
87: \pagerange{\pageref{firstpage}--\pageref{lastpage}} \pubyear{2008}
88: 
89: \maketitle
90: 
91: \label{firstpage}
92: 
93: \begin{abstract}
94: We report results from a systematic study of X-ray emission from black hole 
95: transients in quiescence. In this state mass accretion is thought to follow 
96: the geometry of an outer optically thick, geometrically thin disc and an 
97: inner optically thin, geometrically thick radiatively inefficient accretion 
98: flow (RIAF). The inner flow is likely also coupled to the jets near the
99: black hole that are often seen in such systems. The goal of the study is to 
100: see whether the X-ray emission in the quiescent state is mainly powered by 
101: the accretion flow or the jets. Using data from deep {\it XMM-Newton} 
102: observations of selected black hole transients, we have found that the 
103: quiescent X-ray spectra are, to a high precision, of power-law shape in 
104: the cases of GRO~J1655$-$40 and V404~Cyg. Such spectra deviate significantly 
105: from the expected X-ray spectrum of the RIAF at very low accretion rates. 
106: On the other hand, they can naturally be explained by emission from the 
107: jets, if the emitting electrons follow a power-law spectral distribution 
108: (as is often assumed). The situation remains ambiguous in the case of 
109: XTE~J1550$-$564, due to the relatively poorer quality of the data. We discuss 
110: the implication of the results.
111: \end{abstract}
112: 
113: \begin{keywords}
114: accretion, accretion discs --- black hole physics --- radiation mechanisms: 
115: thermal, non-thermal --  stars: individual (XTE~J1550$-$564, GRO~J1655$-$40, 
116: V404~Cyg) --- X-rays: binaries
117: \end{keywords}
118:        
119: 
120: \section{Introduction}
121: 
122: The majority of X-ray binaries that are known to contain a stellar-mass black 
123: hole are transient X-ray sources. They spend most of their time in the
124: quiescent
125: state, in which the mass accretion rate is thought to be extremely low. 
126: Occasionally, they undergo an outburst during which they may become the 
127: brightest X-ray sources in the sky. The exact mechanism that triggers such
128: an outburst is not entirely understood but is thought to be related to a 
129: sudden surge in the accretion rate that is caused by a thermal instability 
130: in the accretion disc (see reviews, e.g., by King 1995 and Lasota 2001). 
131: During an outburst, the X-ray properties of a black hole transient is often 
132: described empirically in terms of spectral states (e.g., McClintock \& 
133: Remillard 2006; Xue, Wu, \& Cui 2008). 
134: 
135: It is proposed that the spectral states may correspond to different 
136: configurations of the underlying accretion process at different mass 
137: accretion rates (Narayan 1996; Narayan, McClintock \& Yi 1996;
138: Esin, McClintock, \& Narayan 1997). Specifically, when the accretion rate 
139: is high in the high-soft state, the accretion flow is thought to follow the 
140: geometry of the Shakura \& Sunyaev (1973) disk (SSD), which is geometrically 
141: thin and optically thick. This can naturally explain the observed 
142: blackbody-like X-ray spectrum that is characteristic of the high-soft state. 
143: As the accretion rate decreases, the source evolves towards the low-hard 
144: state. In the process, a phase transition is thought to occur in the inner 
145: portion of the disc, in which the accreted matter is heated to nearly local 
146: virial temperatures and may also form an outflow wind (see Narayan 2005 for 
147: a comprehensive review of the models, their evolution, and their applications 
148: to black hole candidates and active galactic nuclei). The accretion flows in
149: this region is a geometrically thick but optically thin configuration, and 
150: is radiatively inefficient. The accretion power is mostly advected into the 
151: black hole or carried away by the outflow (Blandford \& Begelman 1999). Such 
152: a radiatively inefficient accretion flow (RIAF) is 
153: capable of producing hard X-rays by up-scattering ambient soft photons. This 
154: can naturally explain the increasing dominance of the power-law component 
155: of the X-ray spectrum, as the source approaches the low-hard state. The RIAF 
156: model predicts that the trend continues towards even lower accretion rates, 
157: as more of the accretion flow becomes advection dominated, and that the 
158: X-rays in the quiescent state originate entirely from the RIAF (Narayan, 
159: McClintock, \& Yi 1996). 
160: 
161: A more recent development is the realization of the potentially critical role
162: of jets, which seem to be ubiquitous in black hole transients (see review by 
163: Fender 2006 and references therein). Yuan, Cui \& Narayan (2005) demonstrated 
164: that it would be nearly impossible for the RIAF model, which is quite 
165: successful in explaining the X-ray emission from black hole transients in
166: the low-hard state, to 
167: also account for the observed emission at longer wavelengths (radio and IR 
168: in particular). In order to describe the broadband spectral energy 
169: distribution (SED) in the low-hard state, they showed that contributions 
170: from both accretion flow and jets would be needed, with the former 
171: mainly responsible for emission at UV/X-ray wavelengths, the latter for 
172: emission at radio/IR wavelengths, and both for emission in between (cf. 
173: Malzac, Merloni, \& Fabian 2004, who argued that the optical emission might 
174: be dominated by the jets), when 
175: the accretion rates are relatively high. Extrapolating the result of Yuan et 
176: al. (2005) to lower accretion rates, Yuan \& Cui (2005) predicted that the 
177: X-ray emission from the jets would eventually exceed that from the hot flow 
178: (because the former is proportional to the accretion rate $\dot{m}$, 
179: normalized to the Eddington rate, and the latter roughly to $\dot{m}^2$). 
180: In the quiescent state, the X-ray emission should, therefore, be mainly 
181: powered by the jets, at variance with the prediction of the RIAF model. 
182: Falcke, Kording \& Markoff (2004) also postulated that the X-ray emission of 
183: the quiescent state would be dominated by the jets. But, in contrast to 
184: Yuan \& Cui (2005), they argued that this would be the case even for the 
185: low-hard state.
186: 
187: Is it the accretion flow or jets that power the X-ray emission from black hole 
188: transients in the quiescent state? Yuan \& Cui (2005) proposed two 
189: observational tests to answer the question. If the quiescent X-rays are 
190: powered by the jets, they predicted: (1) the radio/X-ray correlation 
191: would steepen when the X-ray flux drops below a characteristic value and 
192: (2) the X-ray spectrum would be of power-law shape. It has been claimed 
193: that the observation of A0620-00 is at odds with the first prediction (Gallo 
194: et al. 2006). However, the conclusion hinged critically on a radio/X-ray 
195: correlation that had been thought to hold for all black hole candidates. 
196: The universality of the radio/X-ray correlation has since been brought into
197: question (Xue \& Cui 2007). The second prediction 
198: is a viable test because the X-ray spectrum of an RIAF would deviate 
199: strongly from power-law shape at sufficiently low accretion rates, when the 
200: density of the flows becomes so low that Comptonization is dominated by 
201: single scattering (Narayan, McClintock, \& Yi 1996; Quataert \& Narayan 1999; 
202: McClintock et al. 2003; Yuan, Cui, \& Narayan 2005).
203: 
204: In this work, we present results from a systematic study of black hole 
205: transients in quiescence. A number of such sources had been observed and 
206: detected 
207: earlier with {\it Chandra} and {\it XMM-Newton} (Kong et al. 2002; Hameury 
208: et al. 2003) but none of the X-ray spectra obtained are of sufficiently 
209: high quality that would allow us to distinguish jet-based and accretion-based 
210: models. To improve the situation, we carried out deep observations of 
211: selected sources with {\it XMM-Newton}. The results reported here are based 
212: on data from these as well as an archival {\it XMM-Newton} observation.
213: 
214: \section{Observations and Data Analysis}
215: 
216: Based on information on the quiescent X-ray fluxes of transient black holes
217: from previous works (Kong et al. 2002; Hameury et al. 2003), we selected
218: V404~Cyg, XTE~J1550$-$564, and GRO~J1655$-$40 for this pilot study. V404~Cyg
219: is the brightest in X-rays among all black hole transients that have been 
220: observed in the quiescent state and is thus the best source for our 
221: investigation. It had already been observed with {\it XMM-Newton} for about
222: 40 ks, with the EPIC detectors in the full-window mode and the medium filter
223: used. We estimated that the quality of the data from that observation would
224: be sufficient for our purposes. For this work, therefore, we simply used the
225: archival data.
226: 
227: XTE~J1550$-$564 had never been observed with {\em XMM-Newton} before, despite 
228: being the first black hole candidate to have its jets directly imaged in 
229: X-rays and also being one of the most active, producing an outburst (of 
230: varying magnitude) every couple of years. The quiescent X-ray luminosity 
231: of XTE~J1550$-$564 is nearly the same as that of V404~Cyg ($\sim 10^{33}$ 
232: $ergs$ $s^{-1}$; Kong et al. 2002) but it lies at a larger distance.
233: It seems to have a harder X-ray spectrum than V404~Cyg but the uncertainty 
234: is quite large (Kong et al. 2002). We observed XTE~J1550$-$564 in quiescence 
235: with {\em XMM-Newton} for about 60 ks in 2007, with the EPIC detectors also 
236: in the full-window mode but the thin filter was used. Unfortunately, the 
237: observation suffered from severe contamination by solar flares. The standard 
238: filtering procedure (as described in the {\it XMM-Newton} Data Analysis 
239: Cookbook\footnote{See http://wave.xray.mpe.mpg.de/xmm/cookbook}) removed 
240: the large flares but smaller ones were still visible after the filtering. 
241: We performed additional filtering, based on the 0.5--10 keV lightcurve of 
242: the whole chip, to minimize the effects of the contamination. The resulted 
243: effective exposure times are only about 
244: 14~ks and 24~ks for the pn and MOS detectors, respectively.
245: 
246: GRO~J1655$-$40 is also relatively luminous in X-rays in the quiescent state
247: but the luminosity seems to vary greatly, from about $3\times 10^{31}$ to 
248: $10^{33}$ ergs/s (Hameury et al. 2003). The source is at roughly the same
249: distance as V404~Cyg. It had already been observed by {\em XMM-Newton} for
250: about
251: 50 ks (Hameury et al. 2003). The X-ray flux measured is about a factor of 5 
252: less than that of XTE~J1550$-$564. Since a similar flux had been obtained 
253: earlier with {\em Chandra} (Kong et al. 2002), this is perhaps the more
254: typical state for the source. The X-ray spectrum of GRO~J1655$-$40 is quite 
255: similar to that of XTE~J1550$-$564 but the spectrum lacks statistics for 
256: our purposes. We observed GRO~J1655$-$40 in quiescence with {\it XMM-Newton} 
257: in 2007 for about 185~ks. The EPIC detectors were operated in the 
258: full-window mode with the thin filter. We should note that this observation 
259: was also partially contaminated by solar flares, but not as severely as in 
260: the case of XTE~J1550-564.
261: 
262: Table~1 summarizes the key characteristics of the {\em XMM-Newton} 
263: observations. We followed the same procedures, as described in the Data 
264: Analysis Cookbook, for all observations in preparing and filtering data, 
265: making light curves, extracting spectra, and generating the corresponding 
266: response matrix files and ancillary response files for subsequent spectral 
267: modeling. Briefly, the source counts were extracted from within a circular 
268: region that is centered on each source and is of radius 40\arcsec\ 
269: (or roughly the 85\% encircled-energy radius of the on-axis PSF), which is 
270: the largest source region that could be used without extending to a different 
271: CCD chip. The background counts were obtained from a different circular 
272: region of radius 80\arcsec\ that is located on the same chip, sufficiently
273: far from any sources.
274: 
275: The raw pn and MOS spectra were re-binned such that each energy bin contains
276: at least 16 counts. They were then filtered to cover the energy range of
277: 0.5--10~keV. A 1\% systematic error was added to the spectra. The initial
278: modeling was carried out in {\it XSPEC} (Arnaud 1996). We chose to jointly
279: fit the individual pn and MOS spectra. For XTE~J1550$-$564, however, we 
280: only used the pn spectrum, because of the poorer statistics of the MOS data, 
281: and limited the energy range to 0.5--7.7~keV, where the signal-to-noise 
282: ratio is adequate. In all three cases, the X-ray spectrum can be fitted 
283: satisfactorily with a simple power law that is attenuated by interstellar 
284: absorption. Figure~1 shows the X-ray spectra of the three sources, along
285: with the best-fit power laws and residuals. The model parameters are 
286: summarized in Table~2. Note that the hydrogen column density was fixed 
287: to the line-of-sight value in the case of XTE~J1550$-$564; the actual 
288: value should be lower but it would hardly affect our results, again due to 
289: the poor quality of the data. Our results on V404~Cyg are in agreement with 
290: those previously obtained from the same dataset (Bradley et al. 2007).
291: 
292: In each case, we used the best-fit power law to compute unabsorbed photon 
293: and energy fluxes for a set of energy bins, to quantify the intrinsic
294: X-ray spectrum of each source. The widths of the bins (typically 
295: 0.2$-$0.3~keV) were chosen to maintain sufficiently small error bars on
296: the fluxes across the entire energy range. The error bars were derived (in 
297: XSPEC) by randomly drawing 1000 sets of parameters and taking the $1~\sigma$ 
298: range of the resulted flux distribution. From the photon and energy fluxes 
299: we also computed the effective energy of each bin ($E_{eff}=F_{en}/F_{ph}$). 
300: The unabsorbed X-ray spectra are shown in Figure~2. It is quite apparent 
301: that the quiescent spectra of GRO~J1655$-$40 and V404~Cyg are, to a high
302: precision, of power-law shape. The situation is a bit ambiguous in the case 
303: of XTE~J1550$-$564, due to the limited statistics of the data.
304: 
305: \section{Modeling Results}
306: 
307: We proceeded to examine the derived spectra with the same coupled 
308: accretion$-$jet model that has provided a good description of the SED of 
309: XTE~J1118+480 over a broad spectral range (from radio to hard X-ray 
310: frequencies; Yuan, Cui, \& Narayan 2005). In this model, the accretion flow 
311: is described as a standard thin disk outside a transition 
312: radius $r_{\rm tr}$ and an RIAF inside this radius. For the quiescent state, 
313: $r_{tr}$ is expected to be very large (Narayan 2005), hence the disc is 
314: cool and likely contributes only to IR/optical emission. The effect of 
315: outflow/convection in the RIAF is taken into account in calculating the 
316: dynamics of the RIAF. The main parameters of RIAF are the viscous parameter 
317: $\alpha$, a parameter describing the strength of the magnetic field $\beta$, 
318: and a parameter $\delta$ which determines the fraction of the turbulent 
319: dissipation which directly heats electrons. For the radiation processes 
320: in the RIAF, we considered synchrotron emission, bremsstrahlung emission,
321: and inverse Compton scattering. Near the black hole, we assumed that 
322: a fraction of the accretion flow is transferred into the vertical 
323: direction to form a jet. Within the jet, internal shocks occur 
324: due to the collision of shells with different velocities. These shocks 
325: accelerate a fraction of the electrons into a power-law energy distribution. 
326: The steady-state energy distribution of the accelerated electrons 
327: is self-consistently determined, taking into account the effect 
328: of radiative cooling. The energy density of accelerated electrons 
329: and amplified magnetic field is determined by two parameters, $\epsilon_e$ and 
330: $\epsilon_B$, which describe the fraction of the shock energy going 
331: into electrons and the magnetic field, respectively. Then the synchrotron 
332: radiation from these accelerated electrons can be calculated. We note that 
333: the effects of inverse Compton scattering of the synchroton photons in the 
334: jet are negligible due to small scattering optical depth. We refer the 
335: readers to Yuan et al. (2005) for a detailed description of the model.
336: 
337: For this work, we fixed all model parameters to the values derived from 
338: fitting the SED of XTE~J1118+480 (Yuan, Cui, \& Narayan 2005), except 
339: for those that are source specific, such as distance, inclination angle, 
340: black hole mass, accretion rate, and $r_{tr}$, which is presumed to depend 
341: on accretion rate. As before, we assumed that the accretion rate is 
342: dependent of radius in the RIAF, $\dot{M}=\dot{M}_{0}(r/r_{tr})^{0.3}$ 
343: (Yuan, Cui, \& Narayan 2005). However, we did {\em not} assume that the 
344: accretion rate is constant in the optically thick disc, because the disc 
345: is not expected to be in a steady state for the quiescent state (e.g., 
346: Lasota 2001). Instead, the accretion rate is expected to decrease with 
347: radius, resulting in a roughly isothermal temperature profile, which 
348: deviates significantly 
349: from the SSD. For this work, therefore, we adopted a constant-temperature 
350: disc for the model (although we also investigated scenarios in which the
351: disc is assumed to be of SSD type). The results of this modeling are also 
352: shown in Fig.~2, with the model parameters summarized in Table~3. 
353: 
354: We also explored the possibility of modeling the X-ray spectra with a pure
355: accretion model (consisting only of disc plus RIAF). The results from our 
356: best attempts are also shown in Fig.~2 (insets); the values of the model 
357: parameters are shown in Table~3. We should note that in the cases of 
358: XTE~J1550-564 and GRO~J1655-40 we had to adopt a lower value for the 
359: viscous parameter in the RIAF (not shown in the table), $\alpha=0.1$ as 
360: opposed to $0.3$ (see Yuan, Cui, \& Narayan 2005), to achieve better fits 
361: to the data.
362: As is apparent from Fig.~2, the high-quality quiescent X-ray spectra of 
363: GRO~J1655$-$40 and V404~Cyg cannot be accounted for by emission from the 
364: RIAF alone. We believe that this conclusion is robust, independent of 
365: theoretical uncertainties associated with the model adopted. Fundamentally, 
366: at very low accretion rates, contribution from multiple scatterings in 
367: the hot flows is negligible, so the Comptonized spectrum deviates strongly
368: from a power law. On the other hand, the observed power-law shape of the 
369: quiescent X-ray spectra can be naturally explained by emission from the 
370: jets, if the spectral energy distribution of the emitting electrons is, 
371: as often assumed, of power-law shape. Therefore, our results are in favor 
372: of a jet origin of the X-ray emission from black hole transients in 
373: quiescence, lending support to the prediction of Yuan \& Cui (2005).
374: 
375: The situation is still ambiguous in the case of XTE~J1550$-$564, due to the 
376: relatively large uncertainties of the measurements. An RIAF origin of the 
377: X-ray emission cannot be ruled out in this case (see Fig.~2). A deeper 
378: {\em XMM-Newton} observation of the source 
379: in quiescence would provide the much needed data to resolve the ambiguity. 
380: In general, we believe that it is important to extend the effort to more
381:  black hole transients. The data would allow us to see whether the 
382: conclusion holds for such systems as a population.
383: 
384: For comparison, we also included in Fig.~2 the published results on the 
385: quiescent IR/optical fluxes of the disc (Casares et al. 1993; Orosz et al. 
386: 2002; Greene et al. 2001). It should, however, be stressed that these 
387: measurements were {\em not} made simultaneously with the X-ray measurements 
388: and should thus be taken with caution, because the sources can vary 
389: significantly even in the quiescent state (e.g., Bradley et al. 2007). 
390: Extrapolating our ``best-fit'' models to the IR/optical wavelength, we 
391: found that the predicted fluxes fall below the values from earlier
392: measurements for XTE~J1550$-$564 and V404~Cyg. We found that the discrepancy
393: was worse when we replaced the constant-temperature disc with the SSD; the
394: real disc is likely to fall in between these two extreme scenarios. The
395: discrepancy could be attributed
396: to the variability of the sources in the quiescent state. For instance, the 
397: accretion rates might have been higher during the IR/optical observations, 
398: so the discs were brighter then. Alternatively, there might be additional 
399: sources of IR/optical emission, for instance, hot spots where gas from the 
400: companion star impacts the accretion disc (cf. McClintock et al. 2003). 
401: We should note, however, that in both cases we were able to find good fits 
402: also to the IR/optical data by adopting different values for the distance 
403: and inclination angle of the systems (that are still within measurement 
404: uncertainties). 
405: 
406: \section{Discussion}
407: 
408: Fender et al. (2003) argued, on the basis of the ``universal radio/X-ray 
409: correlation'', that the energetics of the quiescent state ought to be 
410: dominated by the jets, in the sense that the kinetic power of the jet is 
411: much greater than the X-ray luminosity of the accretion flows. In their 
412: jet-dominated state, however, the X-ray luminosity of the jet is not 
413: necessarily also greater than that of the accretion flows, because the 
414: radiative power of the jet is only of the order of 1\% of the kinetic 
415: power (see Yuan \& Cui 2005 for a more detailed discussion). Here, we have 
416: shown that the quiescent X-rays from transient black hole candidates are 
417: likely to originate from the jets, as opposed to the accretion flows.
418: 
419: In summary, we have, for the first time, found direct evidence that the 
420: quiescent state may be fundamentally different from the low-hard state, 
421: as far as the source of X-ray emission is concerned. Contrary to the view 
422: that the former may be a simple extension of the latter towards lower 
423: accretion rates, our results suggest that the X-ray emission from transient 
424: black holes is dominated by contribution from the jets (or other sources
425: of non-thermal electrons) in the quiescent
426: state, while in the low-hard state it is likely dominated by contribution 
427: from the RIAF (e.g., Esin, McClintock, \& Narayan 1997; Yuan et al. 2007). 
428: 
429: Finally, we would like to emphasize that the prediction of Yuan \& Cui (2005) 
430: is insensitive to the mass of black holes and might thus also hold for 
431: active galactic nuclei (AGN). Wu, Yuan \& Cao (2007) has recently modeled a 
432: sample composed of eight FR I galaxies and found that their X-ray spectra
433: should be dominated by jets, rather than by RIAFs, if their luminosities are 
434: below $\sim 10^{-6}L_{\rm Edd}$, and vice versa, as predicted by Yuan \& Cui 
435: (2005). Wrobel, Terashima, \& Ho (2008) observed two low-luminosity AGN at 
436: 8.5 GHz and found that the observed radio luminosity is within a factor of 
437: 3 of the value that is predicted from the observed X-ray luminosity and the 
438: radio--X-ray--mass relation derived by Yuan \& Cui (2005). 
439: 
440: \section*{Acknowledgments}
441: This research has made use of data obtained through the High Energy 
442: Astrophysics Science Archive Research Center Online Service, provided by 
443: the NASA/Goddard Space Flight Center. The work was supported in part by 
444: NASA through grants NNX07AQ29G and NNX07AH43G and by Natural Science 
445: Foundation of China 
446: through grant 10773024. F.Y. also acknowledges support from the Bairen 
447: Program of Chinese Academy of Sciences.
448: 
449: 
450: \begin{thebibliography}{99}
451: 
452: \bibitem[\protect\citeauthoryear{Arnaud}{1996}]{arn96}
453: Arnaud,~K.~A. 1996, in ASP Conf. Ser. 101, Astronomical
454: Data Analysis Software and Systems V, ed. G. Jacoby \& J. Barnes (San
455: Francisco: ASP), 17
456: \bibitem[\protect\citeauthoryear{Blandford \& Begelman}{1999}]{bb99}
457: Blandford,~R.~D., \& Begelman,~M.~C. 1999, MNRAS, 303, L1
458: \bibitem[\protect\citeauthoryear{Bradley et al.}{2007}]{bradetal07}
459: Bradley,~C.~K. Hynes,~R.~I., Kong,~A.~K., Haswell,~C.~A., 
460: Casares,~J., \& Gallo,~E. 2007, ApJ, 667, 427
461: \bibitem[\protect\citeauthoryear{Casares et al.}{1993}]{cas93} 
462: Casares,~J., Charles,~P.~A., Naylor,~T., \& Pavlenko,~E.~P. 1993,
463: MNRAS, 265, 834
464: \bibitem[\protect\citeauthoryear{Dickey \& Lockman}{1990}]{dicloc90}
465: Dickey,~J.~M. \& Lockman,~F.~J. 1990, ARAA, 28, 215
466: \bibitem[\protect\citeauthoryear{Esin, McClintock, \& Narayan}{1997}]{esin97}
467: Esin A. A., McClintock J. E., Narayan R., 1997, 489, 865
468: \bibitem[\protect\citeauthoryear{Falcke, K\"{o}rding, \& Markoff}{2004}]{falcke04}
469: Falcke,~H., K\"{o}rding,~E., \& Markoff,~S. 2004, A\&A, 414, 895
470: \bibitem[\protect\citeauthoryear{Fender, Gallo \& Jonker}{2003}]{fend03}
471: Fender, R. P.; Gallo, E.; Jonker, P. G. 2003, MNRAS, 343, L99
472: \bibitem[\protect\citeauthoryear{Fender}{2006}]{fend06}
473: Fender,~R.~P. 2006, in Compact Stellar X-ray Sources, eds. 
474: W.~H.~G.~Lewin \& M.~van der Klis (Cambridge: Cambridge University Press), 
475: 381 
476: \bibitem[\protect\citeauthoryear{Gallo et al.}{2006}]{gal06}
477: Gallo,~E, et al. 2006, MNRAS, 370, 1351
478: \bibitem[\protect\citeauthoryear{Greene, Bailyn \& Orosz}{2001}]{gren01}
479: Greene,~J., Bailyn,~C.~D. \& Orosz,~J.~A. 2001, ApJ, 554, 1290
480: \bibitem[\protect\citeauthoryear{Hameury et al.}{2003}]{ham03}
481: Hameury, J.-M., et al. 2003, A\&A, 399, 631
482: \bibitem[\protect\citeauthoryear{Hjellming \& Rupen}{1995}]{hjel95}
483: Hjellming,~R.~M. \& Rupen,~M.~P. 1995, Nature, 375, 464
484: \bibitem[\protect\citeauthoryear{King}{1995}]{king95}
485: King,~A.~R. 1995, in X-ray Binaries, eds. W.~H.~G.~Lewin et al. 
486: (Cambridge: Cambridge Univ. Press), 126
487: \bibitem[\protect\citeauthoryear{Kong et al.}{2002}]{kong02}
488: Kong,~A.~K., McClintock,~J.~E., Garcia,~M.~R., Murray,~S.~S., 
489: \& Barret,~D. 2002, ApJ, 570, 277
490: \bibitem[\protect\citeauthoryear{Lasota}{2001}]{las01}
491: Lasota,~J.-P. 2001, NewAR, 45, 449 
492: \bibitem[\protect\citeauthoryear{Malzac et al.}{2004}]{mal04}
493: Malzac,~J., Merloni,~A., \& Fabian,~A.~C. 2004, MNRAS, 351, 253 
494: \bibitem[\protect\citeauthoryear{McClintock et al.}{2003}]{mcc03}
495: McClintock,~J.~E., Narayan,~R., Garcia,~M.~R., Orosz,~J.~A., 
496: Remillard,~R.~A., \& Murray,~S.~S. 2003, ApJ, 593, 435
497: \bibitem[\protect\citeauthoryear{McClintock \& Remillard}{2006}]{mcrem06}
498: McClintock,~J.~E. \& Remillard,~R.~A. 2006, in Compact Stellar 
499: X-ray Sources, eds. W.~H.~G.~Lewin \& M.~van der Klis (Cambridge: Cambridge 
500: University Press), 157
501: \bibitem[\protect\citeauthoryear{Narayan}{2005}]{nar05}
502: Narayan,~R. 2005, Ap\&SS, 300, 177
503: \bibitem[\protect\citeauthoryear{Narayan}{1996}]{nar96}
504: Narayan,~R. 1996, ApJ, 462, 136
505: \bibitem[\protect\citeauthoryear{Narayan, McClintock \& Yi}{1996}]{narmcyi96}
506: Narayan,~R., McClintock,~J.~E., \& Yi,~I. 1996, ApJ, 457, 821
507: \bibitem[\protect\citeauthoryear{Orosz, Groot \& van~der~Klis}{2002}]{orgvan02}
508: Orosz,~J.~A., Groot,~P.~J., van~der~Klis,~M. et al. 2002, ApJ,
509: 568, 845
510: \bibitem[\protect\citeauthoryear{Quataert \& Narayan}{1999}]{quanar99}
511: Quataert,~E. \& Narayan,~R. 1999, ApJ, 520, 298
512: \bibitem[\protect\citeauthoryear{Shakura \& Sunyaev}{1973}]{shaksuny73}
513: Shakura,~N.~I. \& Sunyaev,~R.~A 1973, A\&A, 24, 337
514: \bibitem[\protect\citeauthoryear{Wrobel, Terashima \& Ho}{2008}]{wrob08}
515: Wrobel,~J.~M., Terashima,~Y., \& Ho,~L.C. 2008, ApJ, 675, 1041
516: \bibitem[\protect\citeauthoryear{Wu, Yuan \& Cao}{2007}]{wu07}
517: Wu,~Q., Yuan,~F., \& Cao,~X. 2007, ApJ, 669, 96
518: \bibitem[\protect\citeauthoryear{Xue \& Cui}{2007}]{xue07}
519: Xue,~Y.~Q., \& Cui,~W. 2007, A\&A, 466, 1053
520: \bibitem[\protect\citeauthoryear{Xue, Wu \& Cui}{2008}]{xue08}
521: Xue,~Y.~Q., Wu,~X.-B., \& Cui,~W. 2008, MNRAS, 384, 440
522: \bibitem[\protect\citeauthoryear{Yuan, Cui \& Narayan}{2005}]{yucunar05}
523: Yuan,~F., Cui,~W. \& Narayan,~R. 2005, ApJ, 620, 905
524: \bibitem[\protect\citeauthoryear{Yuan \& Cui}{2005}]{yuan05}
525: Yuan,~F. \& Cui,~W. 2005, ApJ, 629, 408
526: \bibitem[\protect\citeauthoryear{Yuan et al.}{2007}]{yuan07}
527: Yuan,~F., Zdziarski,~A.~A., Xue,~Y., \& Wu,~X.-B. 2007, ApJ, 
528: 659, 541
529: 
530: \end{thebibliography}
531: 
532: \clearpage
533: 
534: \begin{table}
535: \caption{Summary of Observations}
536: \begin{tabular}{ccccccc} \hline \hline
537: Obs. ID & Source & Date & Duration & \multicolumn{3}{c}{Effective Exposure}  \\
538: & & & (ks) & \multicolumn{3}{c}{(ks)} \\
539: \cline{5-7} \\
540: & & & & pn & MOS1 & MOS2 \\ \hline
541: 0400890201 & GRO~J1655$-$40 & 2007.03.26 & 120 & 54.1 & 75.3 & 73.6  \\
542: 0400890301 & GRO~J1655$-$40 & 2007.02.22 & 65 & 48.8 & 55.3 & 54.0 \\
543: 0400890101 & XTE~J1550$-$564 & 2007.02.25 & 59 & 13.6 & 24.4 & 23.7 \\
544: 0304000201 & V404~Cyg & 2005.11.08 & 40 & 31.0 & 36.8 & 36.4  \\ \hline
545: \end{tabular}
546: \end{table}
547: 
548: \begin{table}
549: \caption{Results of Power-Law Fitting$^1$}
550: \begin{tabular}{lcccc} \hline \hline
551: Source & $N_H$ & $\Gamma$ & $F_{x}$ & $\chi^{2}_{\nu}/dof$ \\
552: & $(10^{21} cm^{-2})$ & & $(10^{-14} erg s^{-1} cm^{-2})$$^b$ & \\ \hline
553: GRO~J1655$-$40 & $6.8^{+1.4}_{-1.1}$ & $2.0^{+0.1}_{-0.2}$ & $5.8^{+0.8}_{-1.8}$ & $1.06/374$ \\
554: XTE~J1550$-$564 & $9$ (fixed)$^a$& $2.1^{+0.6}_{-0.5}$ & $4.9^{+2.1}_{-4.6}$ & $0.96/20$ \\
555: V404~Cyg & $9.4^{+0.8}_{-0.7}$ & $2.11\pm0.09$ & $40^{+4}_{-5}$ & $0.90/285$ \\
556: \hline
557: \end{tabular}
558: \\$^1$ The errors shown represent 90\% confidence intervals for a single 
559: parameter. \\
560: $^a$ Dicky \& Lockman 1990. \\
561: $^b$ Measured X-ray flux in the 0.5--10~keV band. 
562: \end{table}
563: 
564: \begin{table}
565: \caption{RIAF+jet vs. pure RIAF modeling$^1$}
566: \begin{tabular}{lccccccccc} \hline \hline
567: Source&$d^b$&$M_{BH}^b$&$L_{x}/L_{Edd}$&$i^b$&$r_{tr}$&$\dot{m}(r_{tr})$&$\theta_{jet}$&$\dot{m}_{jet}$&$L_{kin}/\dot{M}(10R_{S})c^2$ \\
568: &(kpc)&($M_{\sun}$)&$(10^{-7})$&(\arcdeg)&$(R_S)$&$(10^{-4})$&(\arcdeg)&$(10^{-6})$&$(10^{-3})$ \\ \hline
569: GRO~J1655$-$40 & 3.2 & 6.3 & 1.4 & $70$ & $5000$ & $1.3$ & $85^c$ & $2.35$ & $28$ \\
570: XTE~J1550$-$564 & 5.3 & 9.6 & 2.3 & $73$ & $5000$ & $0.6$ & $73^d$ & $2.4$ & $62$ \\
571: V404~Cyg & 3.5 & 11.7 & 7.0 & $56$ & $5000$ & $5.0$ & $56^d$ & $2.8$ & $8.6$ \\ 
572: \hline
573: GRO~J1655$-$40$^a$ & 3.2 & 6.3 & 1.4 & 70 & 1000 & 1.5 & - & - & - \\
574: XTE~J1550$-$564$^a$ & 5.3 & 9.6 & 2.3 & 73 & 1000 & $1.7$ & - & - & - \\
575: V404~Cyg$^a$ & 3.5 & 11.7 & 7.0 & 56 & 5000 & 18 & - & - & - \\ \hline
576: \end{tabular}
577: \\$^1$ The columns are: source name, distance, BH mass, Eddington scaled
578: luminosity in the 0.5--10~keV band, binary inclination, 
579: RIAF/disc transition radius, accretion rate (in Eddington units) at $r=r_{tr}$,
580: jet inclination, outflow rate (in Eddington units), and jet kinetic 
581: power $\Gamma_{j}(\Gamma_{j}-1)\dot{M}_{jet}c^2$
582: (normalized to accretion power at $r=10 R_S$).  \\
583: $^a$ Pure accretion modeling. \\
584: $^b$ McClintock \& Remillard 2004, and references therein. \\
585: $^c$ Hjellming \& Rupen 1995. \\
586: $^d$ The jet angle is assumed to be equal to the inclination angle of the 
587: binary orbit.
588:  
589: \end{table}
590: 
591: \clearpage
592: 
593: %fig.1
594: \begin{figure}
595: \epsfig{figure=f1a.eps,width=8.5cm,angle=0}
596: \epsfig{figure=f1b.eps,width=8.5cm,angle=0}
597: \epsfig{figure=f1c.eps,width=8.5cm,angle=0}
598: \caption{Power-law fit for GRO~J1655$-$40 (top), XTE~J1550$-$564
599: (middle), and V404~Cyg (bottom). }
600: \end{figure}
601: 
602: %fig.2
603: \begin{figure}
604: \epsfig{figure=f2a.eps,width=8.5cm,angle=0}
605: \epsfig{figure=f2b.eps,width=8.5cm,angle=0}
606: \epsfig{figure=f2c.eps,width=8.5cm,angle=0}
607: \caption{Spectral energy distribution (SED) for GRO~J1655$-$40 (top),
608: XTE~J1550$-$564 (middle), and V404~Cyg (bottom).
609: The solid lines show representative
610: fits with the coupled accretion-jet model (see text). The dashed, dot-dashed, 
611: and dotted lines indicate individual contributions from the hot flows, jets, 
612: and the cool disc, respectively. 
613: For the purpose of illustration, the optical measurements from the literature
614: are included. Since these measurements are not simultaneous with the X-ray
615: measurements, one must exercise caution in drawing conclusions given the
616: potential variability of the sources in the quiescent state.
617: The insets show pure accretion modeling, that is clearly inconsistent
618: with the SED in case of V404~Cyg and GRO~J1655$-$40. The case is ambiguous for
619: XTE~J1550$-$564. }
620: \end{figure}
621: 
622: \end{document}
623: