1: %\documentclass[manuscript]{aastex}
2: %\documentclass[preprint2]{aastex}
3: \documentclass[preprint2]{aastex}
4: \usepackage{amsmath}
5: \usepackage{amssymb}
6: \usepackage{graphicx}
7:
8: %\bibliographystyle{mn2e}
9: \usepackage{natbib}
10: \bibliographystyle{aa}
11:
12:
13: %%% Abbreviations
14: % Common journal abbreviations
15: \def\aap{A\&A} %.......................Astronomy and Astrophysics
16: \def\aaps{A\&AS} %.....................Astronomy and Astrophysics Supplement Series
17: \def\aj{AJ} %..........................Astronomical Journal
18: \def\apj{ApJ} %........................Astrophysical Journal
19: \def\apjl{ApJL} %......................Astrophysical Journal Letters
20: \def\apjs{ApJS} %......................Astrophysical Journal Supplement Series
21: \def\apss{Ap\&SS} %....................Astrophysics and Space Science
22: \def\araa{ARAA} %......................Annual Reviews of Astronomy and Astrophysics
23: \def\josa{J.~Opt.~Soc.~Am.} %..........Journal of the Optical Society of America
24: \def\jpp{J.~Plasma Phys.} %............Journal of Plasma Physics
25: \def\mnras{MNRAS} %....................Monthly Notices of the Royal Astronomical Society
26: \def\nat{Nature} %.....................Nature
27: \def\ptrs{Phil.~Trans.~Roy.~Soc.} %....Philosophical Transactions of the Royal Society
28: \def\physrep{Phys.~Rep.} %.............Physics Reports
29: \def\plb{Phys. Lett. B} %..............Physics Letters B
30: \def\pp{Phys.~Plasma} %................Physical Plasmas
31: \def\prl{Phys.~Rev.~Lett.} %...........Physical Review Letters
32: \def\pr{Phys.~Rev.} %..................Physical Review
33: \def\prd{Phys.~Rev.~D} %...............Physical Review D
34: \def\pre{Phys.~Rev.~E} %...............Physical Review E
35: \def\pasj{Publ.~Astron.~Soc.~Japan} %..Publications of the Astronomical Society of Japan
36: \def\pasp{PASP} %......................Publications of the Astronomical Society of the Pacific
37: \def\prs{Proc.~Roy.~Soc.} %............Proceedings of the Royal Society
38:
39: % Standard abbreviations
40: \def\cf{{cf.}} %.......confer
41: \def\eg{{e.g.}} %......exempli gratia
42: \def\ie{{i.e.}} %......id est
43: \def\etal{{et al.}} %..and others
44: \def\etc{{etc}} %......et cetera
45:
46: % Unit abbreviations
47: %%% Time
48: \def\s{{\rm s}} %...........seconds
49: \def\ns{{\rm n}\s} %........nanoseconds
50: \def\mus{\mu\s} %...........microseconds
51: \def\ms{{\rm m}\s} %........milliseconds
52: \def\ks{{\rm k}\s} %........kiloseconds
53: \def\Ms{{\rm M}\s} %........megaseconds
54: \def\yr{{\rm yr}} %.........years
55: \def\Gyr{{\rm G}\yr} %......gigayears
56: %%% Frequency
57: \def\Hz{{\rm Hz}} %.........Hertz
58: \def\muHz{\mu\Hz} %.........microhertz
59: \def\mHz{{\rm mHz}} %.......millihertz
60: \def\MHz{{\rm MHz}} %.......Megahertz
61: \def\GHz{{\rm GHz}} %.......Gigahertz
62: \def\THz{{\rm THz}} %.......Terahertz
63: %%% Length
64: \def\m{{\rm m}} %...........meters
65: \def\fm{{\rm f}\m} %........femtometers
66: \def\nm{{\rm n}\m} %........nanometers
67: \def\mum{\mu\m} %...........micrometers
68: \def\mm{{\rm m}\m} %........millimeters
69: \def\cm{{\rm c}\m} %........centimeters
70: \def\km{{\rm k}\m} %........kilometers
71: % Angnstroms are already \AA
72: \def\pc{{\rm pc}} %.........parsecs
73: \def\kpc{{\rm k}\pc} %......kiloparsecs
74: \def\Gpc{{\rm G}\pc} %......gigaparsecs
75: \def\au{{\rm au}} %.........astronomical units
76: %%% Mass
77: \def\g{{\rm g}} %...........grams
78: \def\mug{\mu\g} %...........micrograms
79: \def\mg{{\rm m}\g} %........milligrams
80: \def\kg{{\rm k}\g} %........kilograms
81: \def\Ms{M_\odot} %...........solar masses
82: %%% Energy
83: \def\eV{{\rm eV}} %.........electron volts
84: \def\keV{{\rm k}\eV} %......kiloelectron volts
85: \def\MeV{{\rm M}\eV} %......megaelectron volts
86: \def\GeV{{\rm G}\eV} %......gigaelectron volts
87: \def\TeV{{\rm T}\eV} %......teraelectron volts
88: \def\PeV{{\rm P}\eV} %......petaelectron volts
89: \def\XeV{{\rm X}\eV} %......exaelectron volts
90: \def\ZeV{{\rm Z}\eV} %......zetaelectron volts
91: \def\erg{{\rm erg}} %.......ergs
92: \def\J{{\rm J}} %...........joules
93: %%% Magnetic Fields
94: \def\G{{\rm G}} %...........gauss
95: \def\nG{{\rm n}\G} %........nanogauss
96: \def\muG{\mu\G} %...........microgauss
97: \def\mG{{\rm m}\G} %........milligauss
98: \def\kG{{\rm k}\G} %........kilogauss
99: \def\T{{\rm T}} %...........tesla
100: \def\muT{\mu\T} %...........microtesla
101: \def\mT{{\rm m}\T} %........millitesla
102: %%% Angular Resolution
103: % ^\circ is already degrees
104: % ' is alrady minutes
105: % `` is already seconds
106: \def\mas{{\rm mas}} %.......milli-arcseconds
107: \def\muas{\mu{\rm as}} %....micro-arcseconds
108: \def\nas{{\rm nas}} %.......nano-arcseconds
109:
110:
111: \def\d{{\rm d}}
112: \def\e{{\rm e}}
113: \def\SgrA{{Sgr A*}}
114:
115: \title{On The Nature of the Compact Dark Mass at the Galactic Center}
116:
117: \author{Avery E. Broderick\altaffilmark{1} and Ramesh Narayan\altaffilmark{2}}
118: \affil{Institute for Theory and Computation, Harvard-Smithsonian Center
119: for Astrophysics, MS 51, 60 Garden Street, Cambridge, MA 02138, USA}
120: \altaffiltext{1}{abroderick@cfa.harvard.edu}
121: \altaffiltext{2}{rnarayan@cfa.harvard.edu}
122:
123: \shorttitle{The Dark Mass in the Galactic Center}
124: \shortauthors{A.~E. Broderick \& R. Narayan}
125:
126:
127:
128:
129: \begin{document}
130:
131: \begin{abstract}
132:
133: We consider a model in which \SgrA, the $3.7\times10^6\,\Ms$
134: supermassive black hole candidate at the Galactic Center, is a compact
135: object with a thermally emitting surface. For very compact surfaces
136: within the photon orbit, the thermal assumption is likely to be a good
137: approximation because of the large number of rays which are strongly
138: gravitationally lensed back onto the surface. Given the very low
139: quiescent luminosity of {\SgrA} in the near infrared, the existence of
140: a hard surface, even in the limit in which the radius approaches the
141: horizon, places a severe constraint upon the steady mass accretion
142: rate onto the source: $\dot{M}\lesssim10^{-12}\,\Ms\yr^{-1}$. This
143: limit is well below the minimum accretion rate needed to power the
144: observed submillimeter luminosity of \SgrA: $\dot{M} >
145: 10^{-10}\,\Ms\yr^{-1}$. We thus argue that \SgrA~does not have a
146: surface, i.e., it must have an event horizon. The argument could be
147: made more restrictive by an order of magnitude with $\muas$ resolution
148: imaging, \eg, with submillimeter very-long baseline interferometry.
149: \end{abstract}
150:
151: \keywords{Galaxy: center---submillimeter---infrared: general---black hole
152: physics---accretion, accretion disks---gravitational lensing}
153:
154: \maketitle
155:
156:
157: \section{Introduction}
158:
159: Infrared observations of individual stars in the Galactic Center imply
160: the existence of a dark object of mass $M \approx 3.7\times10^6\,\Ms$,
161: constrained to lie within $45\,{\rm AU}$ (or $10^3 GM/c^2$) of the
162: radio source \SgrA~\citep{Scho_et_al:03,Ghez_et_al:05,Eise_et_al:05}.
163: Observations at $3.5\,\mm$ and $7\,\mm$ further constrain the extent
164: of the radio emission from \SgrA~to less than $1-2\,{\rm AU}$ (or
165: $10-20GM/c^2$) \citep{Shen_etal:05,Bowe_etal:04}. The favored
166: interpretation of these observations is that \SgrA~is a supermassive
167: black hole. Indeed, the current constraints rule out many alternative
168: explanations, including clusters of stellar mass compact objects
169: \citep{Maoz:98} and fermion balls \citep{Scho_etal:02}. Nonetheless, it remains to be conclusively
170: demonstrated that the dark mass at the Galactic Center is a true black
171: hole with an event horizon.
172:
173: If {\SgrA} is not a black hole, then it must have a surface at some
174: radius $R$. Although general relativistic considerations coupled with
175: reasonable assumptions on the equation of state of matter require $R
176: \geq 9GM/4c^2$ \citep[see, \eg,][]{Shap-Teuk:86}, alternatives to
177: general relativity exist which allow smaller radii, despite the fact
178: that the exterior spacetime may be arbitrarily close to that predicted
179: by general relativity (\eg, scalar-tensor theories,
180: \citealt{Fuji-Maed:03}, gravastars, \citealt{Mazu-Mott:01}, boson stars
181: \citealt{Torr-Capo-Lamb:00}).
182: Thus, in principle, $R$ could have any value greater than $2GM/c^2$, the horizon
183: radius (we restrict our analysis to non-spinning objects). In this
184: {\it Letter}, we show that current observations do not favor {\SgrA}
185: having such a surface.
186:
187: We assume that any putative surface of {\SgrA} is in steady state in
188: the presence of accreting gas, and that it emits the accreted energy
189: thermally. The latter assumption is reasonable since, even for models
190: in which the radius of the surface approaches $2GM/c^2$, the
191: thermalization timescale is short in comparison to the lifetime of
192: {\SgrA} (or of an observer)\footnote{The thermalization timescale is
193: expected to be on the order of the cooling time of the infalling
194: matter, which is not very different from the free-fall time-scale
195: $\sim 100$ s at the surface. Since the timescale diverges only
196: logarithmically as measured at infinity even for extremely compact
197: configurations (\eg, the gravastar), the thermalization time is
198: increased by only a few orders of magnitude.}. Indeed, for surfaces
199: contained well within the photon orbit, strong gravitational lensing
200: significantly decreases the number of photon trajectories that escape
201: to infinity. Most outgoing rays return to be absorbed by other parts
202: of the surface, so that the object will resemble the classical
203: example of a blackbody: a thermal cavity with a pinhole.
204:
205: If \SgrA~accretes at the Bondi rate from the hot gas surrounding it,
206: the accretion rate is expected to be $\dot M \sim 10^{-6}
207: ~M_\odot\,{\rm yr^{-1}}$ \citep{Baga_etal:03}. A more likely
208: scenario is that the source accretes via a radiatively inefficient
209: accretion flow \citep[RIAF;][and references
210: therein]{Nara-Yi-Maha:95,Yuan-Quat-Nara:03}, with a mass accretion
211: rate in the range $\dot M \sim 10^{-8.5}-10^{-6} ~M_\odot\,{\rm
212: yr^{-1}}$. In a RIAF model, essentially all the potential energy
213: released by the accreting gas would be radiated from the surface of
214: {\SgrA} (assuming it has a surface), with a predicted luminosity at
215: infinity of
216: \begin{equation}
217: L_{\rm surf} \approx \eta \dot M c^2 ,\label{Lsurf}
218: \end{equation}
219: where the efficiency factor $\eta$ is the fraction of the rest mass
220: energy of the infalling gas that is released as radiation. In the
221: Newtonian case, $\eta$ is simply $GM/c^2R$. In general relativity,
222: $\eta$ is given in terms of the gravitational redshift $z$ at the
223: surface:
224: \begin{equation}
225: \eta = z/(1+z), \quad 1+z = (1-2GM/c^2R)^{-1/2}. \label{eta}
226: \end{equation}
227: Although it is highly unlikely that \SgrA~has a radiatively {\it
228: efficient} accretion disk, even such a model requires a fairly large
229: $\dot M$. For example, the observed bolometric luminosity of {\SgrA}
230: of $10^{36} ~{\rm erg\,s^{-1}}$ implies a minimum accretion rate of
231: $\dot M \sim 2\times 10^{-10} ~M_\odot\,{\rm yr^{-1}}$ for a radiative
232: efficiency of 10\%. To within a factor of a few (depending on the
233: nature of the boundary layer at the inner edge of the disk), the
234: luminosity from the surface of {\SgrA} is again predicted to be $\sim
235: \eta\dot Mc^2$ with $\eta$ not very different from (\ref{eta}). All
236: the above estimates are for gas accretion. When stellar capture
237: events are considered, the average accretion rate can be as high as
238: $10^{-5}~M_\odot\,\yr^{-1}$ to $10^{-3}~M_\odot\,\yr^{-1}$, though
239: this would be expected to be in the form of transient accretion events
240: \citep{Mago-Trem:99}. Note that the $\dot M$ estimates given here are
241: from the point of view of a distant observer, i.e., they represent the
242: rate of accretion of rest mass per unit time as measured at infinity.
243:
244: In \S~2, we derive upper limits on $\dot M$ from the observed
245: near-infrared (NIR) fluxes of {\SgrA} and compare these with the
246: various estimates of $\dot M$ given above. On this basis we argue
247: that \SgrA~is unlikely to have a surface and therefore that it must be
248: a black hole. In \S~3, we present theoretical images of the RIAF
249: model discussed in \citet{Yuan-Quat-Nara:03} and
250: \citet{Brod-Loeb:05b}, and show that imaging experiments alone cannot
251: distinguish between a black hole and a compact object with a surface.
252: However, by combining imaging with the argument presented in \S~2, we
253: show that one could make the case for an event horizon stronger. We
254: conclude in \S~4 with a discussion. In what follows, unless otherwise
255: noted, we use geometrized units ($G=c=1$).
256:
257:
258: \section{NIR Limits on the Mass Accretion Rate}
259:
260: The observed flux $F_\nu$ at a frequency $\nu$ from a surface emitting
261: thermal blackbody radiation is simply the blackbody spectrum
262: multiplied by the apparent solid angle of the surface on the sky. For
263: a thermally emitting compact spherical surface at the Galactic Center
264: this is
265: \begin{equation}
266: F_\nu = \frac{2 h \nu^3}{c^2} \frac{\e^{-h\nu/kT}}{1-e^{-h\nu/kT}}
267: \frac{\pi b^2}{D^2}\,, \label{Fnu}
268: \end{equation}
269: where $T$ is the blackbody temperature as measured by the observer (at
270: infinity), $D\simeq8\,\kpc$ is the distance to the Galactic Center,
271: and the apparent size $b$ of the radiating surface is given in terms
272: of its radius $R$ by
273: \begin{equation}
274: b^2 =
275: \left\{
276: \begin{aligned}
277: 27 M^2, && R<3M ,\\
278: R^3/(R-2M), && R\ge3M .
279: \end{aligned}
280: \right.\label{b2}
281: \end{equation}
282: Equation (4) includes the general relativistic correction due to
283: strong lensing. Note that, for $R$ inside the photon orbit $3M$, the
284: apparent size $b$ no longer decreases with decreasing $R$.
285:
286: For a given choice of radius $R$ and a given upper limit on the flux
287: $F_\nu$, equations (\ref{Fnu}) and (\ref{b2}) provide an upper limit
288: $T_{\rm max}$ on the observed temperature (at infinity) and hence a
289: limit on the surface luminosity as measured at infinity: $L_{\rm surf}
290: < 4\pi b^2 \sigma T_{\rm max}^4$. Then, from equations (\ref{Lsurf})
291: and (\ref{eta}) we obtain the following upper limit on the mass
292: accretion rate on the surface,
293: \begin{equation}
294: \dot{M} < 4\pi b^2\sigma T_{\rm max}^4/c^2 \eta \,. \label{Mdot}
295: \end{equation}
296: Thus, for each flux measurement $F_\nu$ of {\SgrA} and an assumed
297: radius $R$, we obtain an estimate of $\dot M_{\rm max}(R)$, the
298: maximum accretion rate for that $R$. This limit, which assumes
299: nothing more than blackbody emission, can be compared with the
300: mass accretion rates which are required to explain the observed
301: luminosity and spectra of {\SgrA} (discussed in \S~1).
302:
303: \begin{deluxetable}{ccc}
304: \tablewidth{0pt}
305: \tablecaption{\label{flux_limits}
306: Near-Infrared Flux Limits of \SgrA}
307: \tablehead{
308: %\colhead{Band} &
309: \colhead{$\lambda\,(\mum)$} &
310: \colhead{$F_\nu\,({\rm mJy})$} &
311: \colhead{Ref.}}
312: \startdata
313: %$H$ &
314: 1.6 & $11$ & \citet{Stol_etal:03}\\
315: %$K$ &
316: 2.1 & $2.8$ & \citet{Ghez_etal:05b}\\
317: %$L$ &
318: 3.8 & $1.28$ & \citet{Ghez_etal:05b}\\
319: %$M$ &
320: 4.8 & $3.5$ & \citet{Clen_etal:04}\\
321: \enddata
322: %\tablenotetext{a}{Dereddened.}
323: \end{deluxetable}
324:
325: \begin{figure}[t!]
326: \begin{center}
327: \includegraphics[width=\columnwidth]{f1.eps}
328: \end{center}
329: \caption{The four curves show upper limits on the mass accretion rate
330: of {\SgrA} as a function of the surface radius $R$, derived from the
331: observed quiescent fluxes at $1.6$, $2.1$, $3.8$ and $4.8\,\mum$
332: listed in Table \ref{flux_limits}. The surface is assumed to radiate
333: as a blackbody. For reference, the photon orbit $R=3M$ is shown by
334: the vertical dashed line. The cross-hatched area at the top
335: corresponds to typical mass accretion rates in RIAF models of \SgrA,
336: and the horizontal dashed line represents the minimum accretion rate
337: needed to power the bolometric luminosity of \SgrA (see \S~1).}
338: \label{mdot_limits}
339: \end{figure}
340:
341: In practice, the strongest limits are placed by observations in the
342: NIR, since the postulated thermal emission from the surface of {\SgrA}
343: peaks in this region of the spectrum. The most constraining
344: observations are listed in Table \ref{flux_limits}, and the
345: corresponding limits on $\dot M(R)$ are plotted in Figure
346: \ref{mdot_limits}. Note that the thermal emission from the surface
347: peaks in the NIR for $R\sim10M$, whereas for larger radii the surface
348: is sufficiently cool that the observed fluxes fall in the Wien regime,
349: resulting in a weakening of the limit on $T_{\rm max}$ and hence on
350: $\dot M$. The difference in character between the limit imposed by
351: observations at $4.8\,\mum$ and the other wavelengths is due to the
352: fact that the former passes into the Rayleigh-Jeans regime at a larger
353: radius. The overall limit on $\dot M$ is simply given by the lower
354: envelope of all four curves in Figure \ref{mdot_limits}; the envelope
355: is dominated by the $3.8\,\mum$ curve.
356:
357: Recent observations at $3.5\,\mm$ and $7\,\mm$ have claimed to resolve
358: {\SgrA}, limiting the radius to $R\lesssim10-20\,M$ at these
359: wavelengths \citep{Shen_etal:05,Bowe_etal:04}. As seen in Figure
360: \ref{mdot_limits}, this limit on the radius, coupled with the NIR flux
361: limits, already restricts the allowed accretion rate in \SgrA~to
362: $\dot{M}\lesssim10^{-12}\,\Ms/\yr$, if the object has a surface.
363: Since the derived limit is two orders of magnitude lower than the
364: lowest $\dot M$ allowed by the observed luminosity of \SgrA ($\sim 2
365: \times 10^{-10} ~M_\odot\,{\rm yr^{-1}}$), and more than three orders
366: of magnitude lower than the lowest accretion rate required by RIAF
367: models (see \S~1), the case for \SgrA~not having a surface, i.e., for
368: it being a black hole, is very strong. As is apparent from Figure
369: \ref{mdot_limits}, observations which can further limit $R$ will place
370: even stronger constraints upon $\dot{M}$, and further strengthen the
371: case for a black hole.
372:
373:
374: \section{Images}
375: \begin{figure*}[!ht]
376: \begin{center}
377: \includegraphics[width=\textwidth]{f2.eps}
378: \end{center}
379: \caption{$350\,\GHz$ images of a RIAF model of {\SgrA} (the $a=0$
380: model in \citealt{Brod-Loeb:05b}) in which the gas accretes onto a
381: compact thermally emitting surface in a Schwarzschild spacetime. The
382: panels correspond to different assumed radii of {\SgrA}, given in the
383: upper left corner of each image, ranging from $1\%$ larger than the
384: horizon to $6M$. For comparison, an image of the same accretion flow
385: onto a black hole is shown in the upper-left panel. All the images
386: correspond to a viewing angle $45^\circ$ above the orbital plane. The
387: brightness scale is normalized separately in each image, ranging from
388: maximum intensity to zero (black). The white circles show the size of
389: the Schwarzschild horizon $2M$. Note that the thermally emitting
390: surface is practically invisible. This is because of its very low
391: temperature ($\sim10^4$ K) compared to the brightness temperature of
392: the relativistic accreting gas ($>10^{10}$ K).}
393: \label{images}
394: \end{figure*}
395:
396: Submillimeter very-long baseline interferometry (VLBI) promises to
397: provide an angular resolution on the order of $\sim20\,\muas$,
398: corresponding to a physical scale $\sim5M$ at the Galactic Center.
399: A significant motivation for developing such a capability is the
400: prospect of imaging the silhouette or ``shadow'' of the central black
401: hole
402: \citep{Falc-Meli-Agol:00,Miyo_etal:04,Doel-Bowe:04,Brod-Loeb:05b,Brod-Loeb:05,Brod-Loeb:05c}.
403: The shadow is due to strong lensing, and its detection would be a
404: strong confirmation of one of the major predictions of general
405: relativity in the limit of strong gravity. Such imaging will also
406: constrain the radius of {\SgrA}'s surface and thus strengthen the
407: argument presented in \S~2.
408:
409: Figure \ref{images} shows theoretical submillimeter VLBI images of a
410: RIAF model of {\SgrA} for a number of assumed radii ranging from $1\%$
411: larger than the horizon to $6M$. In all cases the mass accretion rate
412: is taken to be $10^{-8}\,\Ms/\yr$, the canonical value for RIAF models
413: and orders of magnitude above the limits placed in the preceding
414: section. Surface radii in excess of the photon orbit ($3M$) appear as
415: enlarged silhouettes, as may be seen in the bottom panels. Thus, high
416: resolution imaging can immediately limit the size of any radiating
417: surface in {\SgrA} to less than $3M$. Combined with the NIR flux
418: limits, this would strengthen the limit upon the mass accretion rate
419: by nearly another order of magnitude, making a conclusive case for an
420: event horizon.
421:
422: Interestingly, we see that for radii inside of the photon orbit, the
423: images are nearly indistinguishable from the case in which a horizon
424: is present. This is a direct result of the relatively low temperature
425: of the radiating surface ($\sim 10^4\,{\rm K}$ as measured at
426: infinity) coupled with the fact that at submillimeter wavelengths the
427: emission occurs deep in the Rayleigh-Jeans portion of the spectrum.
428: Thus, while submillimeter imaging can limit the radius of \SgrA, it
429: cannot by itself prove that the object is a black hole. It is only
430: when the imaging results are combined with the NIR flux measurements
431: in the manner discussed in \S~2 that a strong case can be made for the
432: existence of a horizon.
433:
434:
435:
436: \section{Discussion}
437:
438: We have shown in this {\it Letter} that current NIR flux limits on
439: {\SgrA} already place a stringent upper limit upon the mass accretion
440: rate of this compact object, assuming that the object possesses a
441: surface. Comparison of these rates to the observed submillimeter
442: luminosity of the source, and the implied lower limit on the mass
443: accretion rate, leads to a serious contradiction, thus providing
444: strong evidence for the absence of a surface. This argument for
445: {\SgrA} being a black hole is robust \citep[see][ and references
446: therein, for other
447: evidence]{Nara_Gar_McC97,Nara_Gar_McC02,Nara_Heyl02}. The argument
448: applies even when the surface is extremely compact, \eg, as expected
449: in gravastar models \citep[see, \eg,][]{Mazu-Mott:01}. Note from
450: equations (\ref{Lsurf}) and (\ref{eta}) that the extremely large
451: gravitational redshift at the surface of an ultra-compact object
452: ($R\to 2M$) does not cause a reduction in the luminosity observed at
453: infinity. On the contrary, for a given $\dot M$, the observed
454: luminosity is maximum when the redshift goes to infinity \citep[this
455: is in contrast to][in which the intrinsic luminosity measured on the
456: surface, as opposed to the mass accretion rate measured at infinity,
457: was assumed to be fixed]{Abra-Kluz-Laso:02}. In addition, the
458: blackbody assumption on which our argument is based is particularly
459: well motivated when the redshift is large (\S~1).
460:
461: Three critical assumptions underly our conclusions: ({\em i}) the
462: surface is in steady state with respect to the accreting material,
463: ({\em ii}) the surface radiates thermally, and ({\em iii}) general
464: relativity is an appropriate description of gravity external to the
465: surface.
466:
467: As mentioned briefly in \S~1, the assumption of steady state is likely
468: to be a good one, even for surfaces very near the horizon (including
469: those which are separated from the horizon by a Planck length, the
470: minimum scale for which a horizon will not develop). However, it
471: should be noted that for a black hole the unradiated binding energy of
472: the accreting matter contributes to an increase of the black hole's
473: mass. Thus a black hole is an explicit example of an accreting
474: compact object which is not in steady state.
475:
476: Of more concern is the assumption that the surface emits thermally.
477: For models in which large-scale correlations play a significant role
478: (\eg, the gravastar) it is unclear what happens to accreting material.
479: For instance, it is conceivable that one may obtain coherent emission
480: with wavelengths comparable to the correlation length of the surface,
481: which in principle could introduce large deviations from the Planck
482: spectrum. While we cannot rule out such a model, we would like to
483: emphasize that, in general, for the gravastar (or similar) model to
484: remain a viable alternative to the black hole model of {\SgrA}
485: necessarily requires an extremely exotic emission mechanism that
486: deviates enormously from blackbody emission.
487:
488: Finally, some assumption regarding the description of gravity external
489: to the surface is necessary to compute the flux due to a compact
490: surface near a strongly gravitating object. In the absence of a well
491: tested alternative, general relativity is the natural choice and this
492: is what we have selected for our calculations. While we have
493: explicitly considered a non-rotating black hole, we expect rotation
494: (the most obvious extension that one would wish to consider) to
495: primarily broaden the thermal spectrum, changing our results by no
496: more than factors of order unity. Multi-wavelength high-resolution
497: imaging of flares in the Galactic Center has been proposed as a method
498: by which the nature of the spacetime surrounding the Galactic Center's
499: black hole may be probed \citep{Brod-Loeb:05,Brod-Loeb:05c}.
500:
501: To summarize, in the absence of unknown exotic phenomena, the current
502: NIR flux measurements already conclusively imply the existence of an
503: event horizon in the black hole candidate {\SgrA} at the Galactic
504: Center.
505:
506:
507: \acknowledgments We would like to thank Avi Loeb for pointing out the
508: highly enhanced accretion rate of {\SgrA} due to stellar capture
509: events. This work was supported in part by NSF grant AST 0307433.
510: A.E.B. gratefully acknowledges the support of an ITC Fellowship from
511: Harvard College Observatory.
512:
513:
514:
515: \begin{thebibliography}{28}
516: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
517:
518: \bibitem[{{Abramowicz} {et~al.}(2002){Abramowicz}, {Klu{\'z}niak}, \&
519: {Lasota}}]{Abra-Kluz-Laso:02}
520: {Abramowicz}, M.~A., {Klu{\'z}niak}, W., \& {Lasota}, J.-P. 2002, \aap, 396,
521: L31
522:
523: \bibitem[{{Baganoff} {et~al.}(2003){Baganoff}, {Maeda}, {Morris}, {Bautz},
524: {Brandt}, {Cui}, {Doty}, {Feigelson}, {Garmire}, {Pravdo}, {Ricker}, \&
525: {Townsley}}]{Baga_etal:03}
526: {Baganoff}, F.~K., {Maeda}, Y., {Morris}, M., {et~al.} 2003, \apj, 591, 891
527:
528: \bibitem[{{Bower} {et~al.}(2004)}]{Bowe_etal:04}
529: {Bower}, G.~C. {et~al.} 2004, Science, 304, 704
530:
531: \bibitem[{{Broderick} \& {Loeb}(2005{\natexlab{a}})}]{Brod-Loeb:05b}
532: {Broderick}, A.~E. \& {Loeb}, A. 2005{\natexlab{a}}, \apjl, submitted
533: (astro-ph/0508386)
534:
535: \bibitem[{{Broderick} \& {Loeb}(2005{\natexlab{b}})}]{Brod-Loeb:05}
536: {Broderick}, A.~E. \& {Loeb}, A. 2005{\natexlab{b}}, \mnras, 363, 353
537:
538: \bibitem[{{Broderick} \& {Loeb}(2005{\natexlab{c}})}]{Brod-Loeb:05c}
539: {Broderick}, A.~E. \& {Loeb}, A. 2005{\natexlab{c}}, \mnras, submitted
540: (astro-ph/0509237)
541:
542: \bibitem[{{Cl{\'e}net} {et~al.}(2004)}]{Clen_etal:04}
543: {Cl{\'e}net}, Y. {et~al.} 2004, \aap, 424, L21
544:
545: \bibitem[{{Doeleman} \& {Bower}(2004)}]{Doel-Bowe:04}
546: {Doeleman}, S. \& {Bower}, G. 2004, Galactic Center Newsletter, 18, 6
547:
548: \bibitem[{{Eisenhauer} {et~al.}(2005)}]{Eise_et_al:05}
549: {Eisenhauer}, F. {et~al.} 2005, \apj, 628, 246
550:
551: \bibitem[{{Falcke} {et~al.}(2000){Falcke}, {Melia}, \&
552: {Agol}}]{Falc-Meli-Agol:00}
553: {Falcke}, H., {Melia}, F., \& {Agol}, E. 2000, \apjl, 528, L13
554:
555: \bibitem[{{Fujii} \& {Maeda}(2003)}]{Fuji-Maed:03}
556: {Fujii}, Y. \& {Maeda}, K.-I. 2003, {The Scalar-Tensor Theory of Gravitation}
557: (Cambridge : Cambridge University Press, 2003.)
558:
559: \bibitem[{{Ghez} {et~al.}(2005{\natexlab{a}})}]{Ghez_et_al:05}
560: {Ghez}, A.~M. {et~al.} 2005{\natexlab{a}}, \apj, 620, 744
561:
562: \bibitem[{{Ghez} {et~al.}(2005{\natexlab{b}})}]{Ghez_etal:05b}
563: {Ghez}, A.~M. {et~al.} 2005{\natexlab{b}}, astro-ph/0508664
564:
565: \bibitem[{{Magorrian} \& {Tremaine}(1999)}]{Mago-Trem:99}
566: {Magorrian}, J. \& {Tremaine}, S. 1999, \mnras, 309, 447
567:
568: \bibitem[{{Maoz}(1998)}]{Maoz:98}
569: {Maoz}, E. 1998, \apjl, 494, L181
570:
571: \bibitem[{{Mazur} \& {Mottola}(2001)}]{Mazu-Mott:01}
572: {Mazur}, P.~O. \& {Mottola}, E. 2001, gr-qc/0109035
573:
574: \bibitem[{{Miyoshi} {et~al.}(2004){Miyoshi}, {Ishitsuka}, {Kameno}, {Shen}, \&
575: {Horiuchi}}]{Miyo_etal:04}
576: {Miyoshi}, M., {Ishitsuka}, J.~K., {Kameno}, S., {Shen}, Z., \& {Horiuchi}, S.
577: 2004, Progress of Theoretical Physics Supplement, 155, 186
578:
579: \bibitem[{{Narayan} {et~al.}(1997){Narayan}, {Garcia}, \&
580: {McClintock}}]{Nara_Gar_McC97}
581: {Narayan}, R., {Garcia}, M.~R., \& {McClintock}, J.~E. 1997, \apjl, 478, L79
582:
583: \bibitem[{{Narayan} {et~al.}(2002){Narayan}, {Garcia}, \&
584: {McClintock}}]{Nara_Gar_McC02}
585: {Narayan}, R., {Garcia}, M.~R., \& {McClintock}, J.~E. 2002, in The Ninth
586: Marcel Grossmann Meeting, 405--425
587:
588: \bibitem[{{Narayan} \& {Heyl}(2002)}]{Nara_Heyl02}
589: {Narayan}, R. \& {Heyl}, J.~S. 2002, \apjl, 574, L139
590:
591: \bibitem[{{Narayan} {et~al.}(1995){Narayan}, {Yi}, \&
592: {Mahadevan}}]{Nara-Yi-Maha:95}
593: {Narayan}, R., {Yi}, I., \& {Mahadevan}, R. 1995, \nat, 374, 623
594:
595: \bibitem[{{Sch{\" o}del} {et~al.}(2002)}]{Scho_etal:02}
596: {Sch{\" o}del}, R. {et~al.} 2002, \nat, 419, 694
597:
598: \bibitem[{{Sch{\" o}del} {et~al.}(2003)}]{Scho_et_al:03}
599: {Sch{\" o}del}, R. {et~al.} 2003, \apj, 596, 1015
600:
601: \bibitem[{{Shapiro} \& {Teukolsky}(1986)}]{Shap-Teuk:86}
602: {Shapiro}, S.~L. \& {Teukolsky}, S.~A. 1986, {Black Holes, White Dwarfs and
603: Neutron Stars: The Physics of Compact Objects} (New York : Wiley-VCH , 1986.)
604:
605: \bibitem[{{Shen} {et~al.}(2005)}]{Shen_etal:05}
606: {Shen}, Z.-Q. {et~al.} 2005, Nature, 000, 000
607:
608: \bibitem[{{Stolovy} {et~al.}(2003){Stolovy}, {Melia}, {McCarthy}, \&
609: {Yusef-Zadeh}}]{Stol_etal:03}
610: {Stolovy}, S., {Melia}, F., {McCarthy}, D., \& {Yusef-Zadeh}, F. 2003,
611: Astronomische Nachrichten Supplement, 324, 419
612:
613: \bibitem[{{Torres} {et~al.}(2000){Torres}, {Capozziello}, \&
614: {Lambiase}}]{Torr-Capo-Lamb:00}
615: {Torres}, D.~F., {Capozziello}, S., \& {Lambiase}, G. 2000, \prd, 62, 104012
616:
617: \bibitem[{{Yuan} {et~al.}(2003){Yuan}, {Quataert}, \&
618: {Narayan}}]{Yuan-Quat-Nara:03}
619: {Yuan}, F., {Quataert}, E., \& {Narayan}, R. 2003, \apj, 598, 301
620:
621: \end{thebibliography}
622: \end{document}
623: