1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{article}
3: \usepackage{emulateapj}
4: \usepackage{float,epsfig}
5:
6: \newcommand{\re}{$r_e$}
7: \newcommand{\LA}{\mbox{\raisebox{-0.6ex}{$\stackrel{\textstyle<}{\sim}$}}}
8: \newcommand{\GA}{\mbox{\raisebox{-0.6ex}{$\stackrel{\textstyle>}{\sim}$}}}%%%%%%
9: \newcommand{\cxo}{{\sl Chandra}}
10: \newcommand{\xmm}{{\sl XMM/Newton}}
11: \newcommand{\ngc}{{NGC~2403}}
12: \newcommand{\msun}{$M_{\odot}$}
13: \newcommand{\ergl}{erg~s$^{-1}$}
14: \newcommand{\hi}{H{\sc i}}
15: \newcommand{\hii}{H{\sc ii}}
16: \newcommand{\ha}{H$\alpha$}
17: \newcommand{\mdot}{$\dot{M}$}
18: \newcommand{\hst}{{\sl Hubble}}
19: \newcommand{\ros}{{\sl ROSAT}}
20: \newcommand{\cxou}{CXOU~J073650.0+653603}
21: \newcommand{\etal}{et al.}
22: \slugcomment{Submitted to Astrophysical Journal}
23:
24: \begin{document}
25:
26: \title{Discovery of a Transient X-ray Source in the Compact Stellar Nucleus of
27: NGC 2403}
28:
29: \author{
30: Mihoko~Yukita\altaffilmark{1},
31: Douglas~A.~Swartz\altaffilmark{2},
32: Roberto~Soria\altaffilmark{3}, and
33: Allyn~F.~Tennant\altaffilmark{4}%, and
34: %Kajal~K.~Ghosh\altaffilmark{2}
35: }
36: \altaffiltext{1}{University of Alabama in Huntsville, Dept. of Physics,
37: Huntsville, AL, USA}
38: \altaffiltext{2}{Universities Space Research Association,
39: NASA Marshall Space Flight Center, VP62, Huntsville, AL, USA}
40: \altaffiltext{3}{Harvard-Smithsonian Center for Astrophysics,
41: Cambridge, MA, USA}
42: \altaffiltext{4}{Space Science Department,
43: NASA Marshall Space Flight Center, VP62, Huntsville, AL, USA}
44:
45: \begin{abstract}
46: We report the discovery of an X-ray source coincident with the
47: nuclear star cluster at the dynamical center of
48: the nearby late-type spiral galaxy \ngc.
49: The X-ray luminosity of this source varies from below detection
50: levels, $\sim$10$^{35}$~\ergl\ in the $0.5-8.0$~keV band, to
51: 7$\times$10$^{38}$~\ergl\ on timescales between observations
52: of $<$2~months.
53: The X-ray spectrum is well-fit by an accretion disk model
54: consisting of multiple blackbody components and corresponding physically
55: to a compact object mass of \GA 5~\msun.
56: No pulsations nor aperiodic behavior is evident in its X-ray light curve
57: on the short timescales of the individual observations.
58: The X-ray properties of the source are more similar to those of
59: the nuclear source X-8 in M33, believed to be a low-mass X-ray binary,
60: then to those of the low-luminosity active galactic nucleus in NGC~4395.
61: The brightness of the nuclear star cluster, $M_I \sim -11.8$~mag, is typical of
62: clusters in late-type spirals but its effective radius, $r_e \sim 12$~pc,
63: is several times larger than average indicating a relatively relaxed
64: cluster and a low probability of a central massive object.
65: The cluster has a mass \GA10$^{6.5}$~\msun\ and an
66: age of $\sim$1.4~Gyr estimating from its observed colors and brightness.
67: %
68: \end{abstract}
69:
70: \keywords{galaxies: individual (NGC 2403) --- galaxies: nuclei --- galaxies: star clusters --- galaxies: evolution --- X-rays: galaxies --- X-rays: binaries}
71:
72: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
73: \section{Introduction}
74:
75: Tight correlations have been measured between
76: the masses of central supermassive black holes in early-type galaxies
77: and their bulge mass (Magorrian \etal\ 1998; H\"{a}ring \& Rix 2004),
78: luminosity (Kormendy \& Richstone 1995),
79: central velocity dispersion (Ferrarese \& Merritt 2000; Gebhardt \etal\ 2000;
80: Tremaine \etal\ 2002),
81: and central light concentration (Graham \etal\ 2001).
82: It is unclear if these correlations extend to less massive galaxies and
83: to galaxies of later morphological type.
84: If so, do the scaling properties still hold? That is, do disk-dominated
85: late-type spirals with small bulges host intermediate-mass black holes
86: or is there a minimum galaxy mass below which
87: black holes fail to form and/or to grow?
88:
89: High-resolution \hst\ images (Carollo \etal\ 1998; B\"{o}ker \etal\ 2002, 2004)
90: show that the majority, $\sim$75\%, of late-type spirals have distinct compact
91: stellar nuclei or nuclear star clusters (NSCs).
92: These are not, however, a smooth extension from the massive bulges
93: of early-type galaxies.
94: They are much more dense and more compact (Walcher \etal\ 2005);
95: hence, while typical NSC masses can be comparable to those of
96: dwarf spheroid galaxies, they are 4 orders of magnitude denser.
97: In this sense, NSCs are more like
98: Milky Way globular clusters with scaled-up masses and densities.
99: In addition, they are much more luminous than the old globular clusters
100: (B\"{o}ker \etal\ 2004) due in large part to the comparatively young age,
101: 100~Myr, of their most luminous stellar component (Rossa \etal\ 2006).
102: Still, their dynamical masses
103: suggest a hidden older population is present as well (Walcher \etal\ 2005;
104: Rossa \etal\ 2006) so that they may grow through occasional bursts of
105: star formation and may be
106: forming stars in the current epoch at low levels.
107:
108: While the NSCs are, by definition, at or very near the dynamical centers
109: of their host galaxies (e.g. B\"{o}ker \etal\ 2002), it is not certain
110: if they originated there nor whether or not they are presently
111: accreting gas and forming stars.
112: It has been suggested that
113: some of the most massive globular clusters may be relic dE nuclei tidally
114: stripped and absorbed into larger galaxies (e.g. Freeman 1993;
115: Layden \& Sarajendini 2000).
116: If they commonly host (intermediate-mass) black holes, then
117: they may be potential candidates for ultraluminous X-ray sources
118: (King \& Dehnen 2005) and, if they migrate to the cores of the larger
119: host galaxies, may be the seed (intermediate-mass) black holes that evolve
120: into supermassive black holes.
121:
122: The SAB(s)cd galaxy
123: \ngc\ is a member of the M81 group of galaxies at a distance
124: $D=3.2$~Mpc (1\arcmin$=$1~kpc; Madore \& Freedman 1991).
125: \ngc\ lacks a central bulge
126: but does host a luminous compact NSC
127: (Davidge \& Courteau 2002).
128: We derive physical properties for this cluster from recent \hst\
129: Advanced Camera for Surveys observations and other archival data in
130: \S~\ref{s:cluster}.
131: We have discovered a bright transient X-ray source within the
132: \ngc\ NSC in archival \cxo\ and \xmm\ observations.
133: The nuclear transient is easily resolved from nearby X-ray sources
134: including the well-known ultraluminous X-ray source located about
135: 2.\arcmin 6 to the west of the nucleus (e.g., Swartz \etal\ 2004).
136: We present the X-ray spectra and light curve of the nuclear source
137: in \S~\ref{s:Xtrans}.
138: %
139: The optical properties of the NSC and the X-ray properties
140: of the transient source are compared to other nearby compact nuclei
141: in \S~\ref{s:compare}.
142: We find the NSC is older and less compact than typical for late-type galaxies.
143: The properties of the X-ray source are consistent with
144: an X-ray binary containing an $\sim$5~\msun\ compact object or larger
145: accreting from a low-mass companion.
146: Further discussion is given in \S~\ref{s:discuss}
147:
148: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
149: \section{The Nuclear Star Cluster in \ngc} \label{s:cluster}
150:
151: The contrast between the nucleus of \ngc\ and the underlying
152: galaxy disk is clearly visible in optical and near-IR images where
153: it appears extended at high resolution (Figure~\ref{f:opt_image}).
154: We used the method described in B\"{o}ker \etal\ (2004) to analyze the morphology
155: of the cluster.
156:
157: Calibrated \hst\ images acquired with the Advanced Camera for Surveys (ACS)
158: operated in WFC mode were obtained from the Multimission Archive at
159: STScI\footnote{http://archive.stsci.edu/}.
160: These images were already cleaned of cosmic rays and bad pixels and were
161: corrected for geometric distortion;
162: we performed no further processing of the data.
163: The data were taken 2004-08-17 and include images through the F475W, F606W,
164: F658N, and F814W filters (dataset identifiers J90ZX1010 through 1040).
165:
166: \begin{center}
167: \includegraphics[angle=0,width=\columnwidth]{f1.eps}
168: \figcaption{3-color combined \hst/ACS image of the 20\arcsec$\times$20\arcsec\ region centered on the nuclear star cluster in \ngc. Colors blue, green, and red correspond to
169: the filters F475W, F606W and F814W, respectively. The image has been
170: rotated for display purposes such that north is up and east is to the left.
171: \label{f:opt_image} }
172: \end{center}
173:
174: The 200\arcsec$\times$200\arcsec\ \hst/ACS field contained both
175: SN~2004dj (the target of the original investigation) and the nucleus of \ngc.
176: Although the location of SN~2004dj is known to high precision
177: (Beswick \etal\ 2005) and the ACS resolution is $\sim$0.\arcsec 1 (with 0.\arcsec 05~pixel$^{-1}$ sampling), the supernova is saturated in these
178: images and this is the dominant contribution to the uncertainty in the
179: position of the nucleus:
180: Our best estimated position for the nucleus is
181: R.A.$=$7$^{\rm h}$36$^{\rm m}$50.$^{\rm s}$070,
182: Decl.$=+$65$^{\circ}$36$^{\prime}$3.$^{\prime\prime}$54 (J2000.0)
183: with an uncertainty of $\sim$0.\arcsec 1 radius
184: based on a circular Gaussian model fit to the location of the supernova
185: in the F606W filter image.
186: The position of the nucleus is consistent with the kinematic center of the galaxy
187: determined by Fraternali \etal\ (2002) from the \hi\ rotation curve.
188: % R.A.$=$7$^{\rm h}$36$^{\rm m}$50.$^{\rm s}$66$\pm$1.48,
189: % Decl.$=+$65$^{\circ}$36$^{\prime}$2.$^{\prime\prime}$2$\pm$5.5.
190:
191: A model point spread function (PSF) of the \hst/ASC detector at the location of the source was constructed using TinyTim (Krist \& Hook 1997) assuming a spectral energy distribution equivalent to that of an A5V star (following B\"{o}ker \etal\ 2004). We then fit analytic models to the surface brightness distribution using the program ISHAPE (Larsen 1999) which convolves the model PSF with these analytic profiles.
192: ISHAPE\footnote{ISHAPE documentation is available from http://www.astro.uu.nl/$\sim$larsen/baolab/}
193: allows for circular or elliptical models but cannot reproduce the
194: azimuthal asymmetries clearly seen in the \hst\ images of \ngc.
195: Therefore, we restricted our selection to circular
196: Moffat and King model profiles.
197: The standard power indices and concentration parameters for these models
198: (see Larsen 1999) provided poor fits to the data.
199: A trial King model applied to the azimuthally-averaged radial profile
200: suggested a Moffat model with power index of 1.35 would improve the model
201: fit. This did give the best fit of all our trials.
202: Figure~\ref{f:ishape} displays this model, the data, and the fit residuals.
203: Following B\"{o}ker \etal\ (2004), we derive an effective
204: radius of $r_e = 11.7\pm0.1$~pc from the ISHAPE fitting.
205: This is larger than the average, $\sim$3.5~pc, found by
206: B\"{o}ker \etal\ (2004) for a sample of 39 NSCs in late-type galaxies.
207:
208: %King model fit to radial profile is acceptable fit 35.2/100 radial bins
209: %rc=4.54$\pm$0.35 index=1.22$\pm$0.06
210: %King is $\propto (1+r/rc)^2)^{-index}$
211: %fixing index to 3/2 gives $\chi^2=77.8$ and rc=6.0
212:
213: We estimate the intrinsic, background-subtracted, $I$-band luminosity to be
214: 1.1$\times$10$^{39}$~\ergl\ in a $2\pi r_e^2$ region centered on the NSC
215: after correcting for an extinction of $E(B-V)=0.2$~mag (see below).
216: The average surface brightness,
217: $I_e=323$~$L_{\odot}$~erg~s$^{-1}$~pc$^{-2}$, is near the low luminosity
218: range of the NSCs
219: in Sd galaxies in the B\"{o}ker \etal\ (2004) sample and near the high
220: luminosity end of Milky Way globular clusters (cf. their Fig.~5).
221: The observed absolute blue magnitude is M$_{B} = -9.3$~mag
222: (giving an extinction-corrected M$_B^{\rm o} = -10.2$~mag, see below).
223:
224: \begin{center}
225: \includegraphics[angle=0,width=\columnwidth]{f2.eps}
226: \figcaption{ISHAPE (Larsen 1999) model fit to the $I$-band surface brightness distribution of the nuclear star cluster in \ngc. {\sl From left to right}: Model, data, and fit residuals. Grayscale of model and data cover the same range. The model is a Moffat profile with a power index of 1.35 (see text).
227: \label{f:ishape}}
228: \end{center}
229:
230: Lacking an optical spectrum of the \ngc\ nucleus
231: we cannot accurately constrain the mass and age of the cluster.
232: However, we can obtain estimates of the cluster age from the observed
233: colors and the cluster mass from M$_B^{\rm o}$
234: using the stellar synthesis models of, e.g., Leitherer \etal\ (1999).
235: The observed colors are (using the conversions from \hst/ACS to
236: Johnson-Cousins given by Sirianni \etal\ 2005)
237: $B-V = 1.0$~mag, $V-I = 1.2$~mag,
238: and $J-K = 0.6$~mag (the latter from Davidge \& Courteau 2002).
239: These colors can be self-consistently reproduced by a single starburst
240: episode aged $\sim$1.4~Gyr viewed through an extinction of $E(B-V)= 0.2$~mag.
241: They cannot be reproduced by models with continuous star formation nor can
242: the metallicity of the cluster be radically different from the solar value.
243: The implied intrinsic colors are consistent with those of G2V to K0V stars.
244: As shown in \S~\ref{s:x_spectrum}, the extinction estimate is consistent with
245: the hydrogen column density towards the NSC X-ray source according to
246: the best-fitting X-ray spectral model.
247:
248: Rossa \etal\ (2006) show there is a strong correlation between the observed
249: $B-V$ color of NSCs and their luminosity-weighted cluster age, $\tau_{\rm L}$.
250: Using their linear correlation coefficients gives
251: $\tau_{\rm L} = 10^{9.8 \pm 0.4}$~yr.
252: The age-mass correlation (Rossa \etal\ 2006) then gives
253: $\log M = 10^{8.3\pm1.8}$~\msun\ for the total mass of the NSC.
254: These values are much higher than expected for a late-type spiral galaxy
255: like \ngc\
256: based on the (weaker) correlations Rossa \etal\ (2006) find
257: between cluster mass or cluster age and galaxy morphological type.
258:
259: There is no correction for extinction in the work of Rossa \etal\ (2006).
260: One possibility is that the NSC in \ngc\ is more reddened than typical.
261: Figure~\ref{f:opt_image} shows the cluster is, in fact, at the edge of a dark
262: region however our best-fitting extinction is only $E(B-V)= 0.2$~mag.
263: Using the implied intrinsic $B-V = 0.8$~mag color reduces the estimated age to
264: $\tau_{\rm L} = 10^{9.3 \pm 0.3}$~yr and the cluster mass to
265: $M = 10^{7.6\pm1.7}$~\msun.
266: This age is consistent with the age estimated above from the colors using
267: the starburst models of Leitherer \etal\ (1999).
268: A mass estimate can also be deduced from the age and intrinsic blue
269: magnitude using these starburst models.
270: The mass estimated in this way is $M \sim 10^{6.4}$~\msun\
271: depending weakly on starburst metallicity and IMF.
272:
273: The age estimated here is much older than the $\sim$100~Myr age deduced by
274: Davidge \& Courteau (2002) for individual AGB stars in the vicinity
275: of the nucleus.
276: (Many of these stars are visible as blue sources in the composite \hst\ image
277: of Figure~1.)
278: Davidge \& Courteau (2002) point out that the $J-K$ color of the cluster
279: is somewhat bluer than these surroundings suggesting an age gradient.
280: According to Leitherer \etal\ (1999), the $J-K$ color for a starburst
281: becomes bluer with age (after an early reddening phase)
282: implying that the NSC in \ngc\ is older than its surroundings.
283: This is consistent with our calculations.
284:
285: Although the bulk of the star formation in the cluster appears to have
286: occurred some $\sim$1.4~Gyr ago,
287: there may be a low level of current star formation present
288: (and, likely, an underlying older population of stars as well).
289: For star-forming clusters, the
290: current star formation rate can be estimated from the \ha\ luminosity.
291: We have constructed a continuum-subtracted \ha\ image from the \hst/ACS
292: images and find no net \ha\ emission;
293: the 2$\sigma$ upper limit to the \ha\ luminosity
294: within a 1.\arcsec 5 radius circle about the NSC is 3$\times$10$^{35}$~\ergl.
295: Thus there is no evidence for current star formation in the NSC.
296:
297: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
298: \section{The Transient X-ray Source in the Nucleus of \ngc } \label{s:Xtrans}
299:
300: The nuclear region of
301: \ngc\ was observed with \cxo\ ACIS-S four times over a 3.7 year interval
302: and three times with \xmm\ within this same time interval.
303: Table~1 provides a log of these observations.
304: Earlier observations with the {\sl Einstein}, {\sl ASCA}, and \ros\
305: Observatories show no evidence for a nuclear X-ray source although the
306: quality of the data is poor compared to the recent \cxo\ and \xmm\
307: observations.
308:
309: We obtained level 1 event lists from the \cxo\ data
310: archive\footnote{http://cxc.harvard.edu/cda/} and reprocessed them using the Chandra Interactive Analysis of Observations (CIAO) version 3.3.0.1 tool {\tt acis\_process\_events} and the calibration database (CALDB) version 3.2.1. Reprocessing removed pixel randomization and applied CTI and
311: time-dependent-gain corrections. We then filtered the data of events with non-{\sl ASCA} grades and bad status bits as well as hot pixels and columns.
312:
313: The European Photon Imaging Camera (EPIC) event lists from the \xmm\ observations were obtained from the HEASARC data
314: archive\footnote{http://heasarc.gsfc.nasa.gov/db-perl/W3Browse/w3browse.pl}.
315: We processed the event lists using the current calibration files with the routines in SAS 6.5.0.
316: %The ARFs and RMFs were constructed using
317: %the SAS tasks {\tt arfgen} and {\tt rmfgen}, respectively.
318:
319: We checked for periods of high background that could affect our spectra and timing analysis; excluding some short intervals from the \xmm\ datasets. The final Good Time Intervals for data used in our analysis are listed in Table~1.
320:
321: \subsection{X-ray Light Curve}
322:
323: A bright, point-like X-ray source was detected in several of the observations
324: at the location of the nuclear star cluster.
325: We used the common source, SN~2004dj, to match the location of the X-ray
326: source and the NSC. The X-ray source is well within the $>$1\arcsec\
327: optical extent of the cluster.
328: Figure~\ref{f:X_lc} shows the 0.5~--~8.0~keV light curve of the source,
329: designated \cxou, constructed from the average flux during
330: each individual observation based on spectral fits (\S~\ref{s:x_spectrum}).
331: For observations in which the source was not detected,
332: a 2$\sigma$ flux upper limit was calculated by scaling from the fitted
333: measurements by the background-subtracted count rate in an appropriate
334: source region.
335: The corresponding observed X-ray luminosities, assuming a distance of 3.2~Mpc,
336: are tabulated in Table~1.
337: The X-ray source was undetected in five observations and orders of magnitude above (\cxo) detection thresholds in three observations which qualifies
338: it as a transient source (e.g. van~Paradijs \& McClintock 1995).
339: The shortest measured interval between a detection and a non-detection is
340: 21~days (observations number 5 and 6 of Table 1) but the \xmm\ upper limit
341: is not particularly compelling. The time between the \cxo\ non-detection
342: observation number 4 and the detection observation 6 is 41~days.
343:
344: We also inspected the light curve from observations in which the source
345: was detected for variability during the observation.
346: These X-ray light curves, binned into 1~ks intervals, displayed no
347: conspicuous variability and were formally consistent with a constant flux
348: model.
349: Searches for short-term aperiodic variability using
350: a power-spectrum analysis found no excess power above the Poisson noise.
351: Searches for coherent pulsations, using the $Z_n^2$ statistics (Buccheri
352: \etal\ 1983), detected nothing significant.
353:
354: \begin{center}
355: \includegraphics[angle=-90,width=\columnwidth]{f3.eps}
356: \vspace{10pt}
357: \figcaption{Lightcurve of the X-ray source in the nucleus of \ngc.
358: Time is measured since the first \cxo\ observation on 2001-04-17.
359: Upper limits for non-detections are represented with arrows.
360: Crosses denote \cxo\ observations.
361: Triangles denote \xmm\ observations. Error bars denote two-sided errors.
362: Table~1 provides the numerical values
363: shown here.
364: \label{f:X_lc}}
365: \end{center}
366:
367: \subsection{X-ray Spectrum} \label{s:x_spectrum}
368:
369: Models were fit to observed spectra of the source.
370: Table~2 lists the models attempted, the resulting best-fit model parameters
371: with 90\% confidence extremes for a single interesting parameter,
372: and the fit statistic for each of the three observations in which the source
373: was detected.
374: \cxou\ was most luminous during the \xmm\ observation of 2003 September.
375: The spectrum from this observation is shown in Figure~\ref{f:X_xmmSpec}.
376: %
377: Following the procedures outlined
378: in Page \etal\ (2003),
379: the average \xmm\ spectrum from the three detectors was modeled in XSPEC
380: using the $\chi^2$ fit statistic.
381: Absorbed {\tt diskbb} and {\tt powerlaw} models were applied.
382: Adding additional model components
383: did not significantly improve the fit statistics according to the F-test.
384: We used an unbinned spectrum with C-statistic in XSPEC in fitting
385: the low-count \cxo\ observations.
386: Only single-component models were applied to the \cxo\ spectra.
387: It was necessary to fix the
388: multiplicative absorption model component at the
389: Galactic value $n_H = 4 \times 10^{20}$~cm$^{-2}$ in fits to some of the \cxo\
390: data to prevent convergence to an unphysically low value.
391:
392: The best-fitting model in all cases is the {\tt diskbb} model
393: (Makishima \etal\ 1986, 2000) representing a spectrum dominated by a
394: geometrically-thin accretion disk which extends to the innermost
395: stable orbit often referred to as a high/soft state
396: (Remillard \& McClintock 2006).
397: The model fit parameters are proportional to the inner
398: disk temperature and radius or, alternatively,
399: can be expressed in terms of the
400: mass of the central object.
401: These parameters are listed in Table~2. They suggest a compact
402: object mass, $M \sim 2.3/T_o^2 (L_{\rm disk,bol}/10^{38})^{1/2}$
403: where $T_o \sim T_{\rm in}$ is the disk color temperature,
404: of about 5~\msun\ and that the peak observed luminosity is
405: therefore about the Eddington value.
406: The mass could be up to about a factor-of-two larger depending on viewing
407: angle, black hole spin, and hardening factor.
408:
409: \begin{center}
410: \includegraphics[angle=-90,width=\columnwidth]{f4.eps}
411: \vspace{10pt}
412: \figcaption{\xmm\ spectrum of \cxou\ with the best fit disk-blackbody model
413: ({\sl upper panel}) and fit residuals.
414: \label{f:X_xmmSpec}}
415: \end{center}
416:
417: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
418: \section{Comparison to Other Nuclear Star Clusters} \label{s:compare}
419:
420: As mentioned in the Introduction, some 75\% of late-type spirals host
421: NSCs.
422: In this section we briefly compare \ngc\ to two well-studied objects
423: with very different interpretations for their X-ray emission mechanisms:
424: NGC~598 (M33),
425: another Scd galaxy with a NSC and a bright point-like nuclear X-ray source,
426: is thought to contain a luminous X-ray binary; and
427: NGC~4395, the optically least-luminous broad-line AGN known,
428: is thought to contain a ``proper'' nuclear black hole at the low end of
429: their mass distribution ($M \sim 10^5$~\msun).
430:
431: \subsection{M33 X8}
432:
433: NGC~598 is a disk-dominated late type spiral galaxy as is \ngc.
434: Its NSC is as luminous ($M_B=-10.2$, Kormendy \& McClure 1993) as that of \ngc\
435: but much more compact (FWHM $<$1 pc, Gordon \etal\ 1999).
436: Analysis of the cluster optical spectrum shows it
437: formed from two major starburst episodes; one $\sim$1~Gyr ago and
438: the other occurring only 40$-$70~Myr in the past (O'Connell 1983;
439: Gordon \etal\ 1999; Long \etal\ 2002).
440: %
441: The best estimated upper limit for the mass of any central compact object
442: in the cluster is $M_{\rm BH}$$\sim$1500~\msun\ based on modeling the narrow-slit
443: \hst\ spectrum (Gebhardt \etal\ 2001).
444:
445: The nucleus contains a bright unresolved X-ray source,
446: designated X-8 (Trinchieri \etal\ 1988), with a luminosity as high as
447: $\sim$2$\times$10$^{39}$~\ergl\ in the 0.5$-$10~keV band
448: (Foschini \etal\ 2004).
449: It has long been debated whether this source could be a weak AGN.
450: The optical spectrum shows no evidence of AGN activity although a
451: strong component would be expected if the measured X-ray flux is extrapolated
452: using the known X-ray to optical flux relationship for AGNs
453: (Long \etal\ 2002).
454: The current consensus is that X-8 is a luminous
455: X-ray binary in the NSC of NGC~598
456: rather than a weak AGN (e.g., La~Parola \etal\ 2003).
457: The mass of the compact object must be $\sim$10~\msun\ or more unless it is
458: radiating above the Eddington limit in the X-ray band.
459:
460: The source X-ray flux has remained at about its current level since discovery
461: although it does vary by about a factor of two on short timescales.
462: This is in contrast with many low-mass XRBs which are soft X-ray transients that
463: vary by factors of 100 or more on timescales of weeks or months (as does the
464: nuclear source in \ngc).
465: It has been suggested that M33 X-8 is similar to GRS~1915$+$105
466: (e.g., Long \etal\ 2002; Dubus \& Rutledge 2002)
467: in that both are steady sources with comparable X-ray luminosities
468: and spectral states.
469: Both are also radio sources (Dubus \& Rutledge 2002).
470: GRS~1915$+$105 is a low-mass X-ray binary with a 14$\pm$4~\msun\ compact
471: object accreting from a low mass, evolved (K or M giant) companion
472: (Greiner, Cuby, \& McCaughrean 2001).
473:
474: \subsection{NGC 4395}
475:
476: NGC~4395 is the optically least-luminous AGN known.
477: It is a SA(s)m dwarf galaxy with a distinct though weak type 1 Seyfert spectrum
478: (Filippenko \& Sargent 1989).
479: The nuclear brightness is only $M_B \sim -11$~mag which makes it only
480: twice as bright as the NSC in \ngc.
481: Filippenko \& Ho (2003) estimate the mass of its compact object to be less than
482: $\sim$10$^5$~\msun.
483: Its X-ray spectrum is a moderately absorbed,
484: $N_H$$=$(1.2$-$2.3)$\times$10$^{22}$~cm$^{-2}$, hard power law
485: ($\Gamma$ \LA 1.5, Shih \etal\ 2003; $\Gamma\sim0.6$, Moran \etal\ 2005).
486: The absorption-corrected 2$-$10~keV luminosity is
487: nearly 10$^{40}$~\ergl\ (Iwasawa \etal\ 2000; Moran \etal\ 2005).
488: It is highly variable in X-rays (Moran \etal\ 1999, 2005; Shih \etal\ 2003) exhibiting
489: factor of 10 changes in X-ray flux in less than 2~ks.
490:
491: The low power law index, moderate absorption, and rapid flux variability
492: of the AGN in NGC~4395 is unlike the behavior of the nuclear source in \ngc\
493: in all respects.
494:
495: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
496: \section{Discussion} \label{s:discuss}
497:
498:
499: We have presented a new study of the NSC in \ngc.
500: We discovered that it contains a transient X-ray source;
501: its X-ray luminosity varies from $\sim$7$\times$10$^{38}$ to
502: \LA 5$\times$10$^{34}$ \ergl\
503: over a few months. The question is whether this accreting source
504: is the galaxy's nuclear black hole or, instead, a lower-mass
505: black hole formed in the star cluster from stellar processes.
506:
507: Its luminosity is consistent both with a stellar-mass black hole (BH)
508: and with a supermassive black hole (SMBH) emitting a few orders of magnitude
509: below its Eddington limit (as are the large majority
510: of SMBHs in the Local Universe). Its X-ray variability
511: pattern, over at least 4 orders of magnitude over
512: a few months to years, is more consistent with typical
513: X-ray binaries. Its X-ray spectrum when the source is bright
514: is exactly what we expect from
515: a BH of mass $\sim$5$-$10~\msun\ in the thermal dominant
516: state ($L_{\rm X} \sim 8\times10^{38}$~\ergl\ $ \sim L_{\rm Edd}$,
517: $T_{\rm in}\sim 1$~keV,
518: $R_{\rm in}(\cos \theta)^{1/2} \sim 50$~km). SMBHs may have comparable
519: X-ray luminosities but tend to have power-law spectra
520: in the X-ray band, because their characteristic disk
521: temperatures are \LA 0.1 keV. We conclude that the X-ray
522: source in the NSC of \ngc\ is more likely to be
523: an ordinary X-ray binary, similar to the nuclear
524: source in the NSC of M33.
525:
526: X-ray studies of nuclear sources associated with NSCs
527: provide an additional tool to understand and quantify
528: the relation between the formation of galactic bulges
529: and the presence and properties of a central ``compact
530: massive object'' (CMO). Ideally, one should look for correlations
531: that extend unbroken from large ellipticals to small
532: late-type disk galaxies. One such fundamental
533: correlation seems to hold between the bulge mass
534: and the CMO mass (Ferrarese \etal\ 2006; Wehner \& Harris 2006), where
535: the CMO is a supermassive BH in ellipticals
536: and early-type spirals and a NSC in late-type spirals.
537: The CMO mass is $\sim$1/500 of the spheroidal mass
538: (bulge mass in spirals). Qualitatively, this suggests
539: that the initial galaxy assembly processes led to
540: the rapid collapse of a nuclear gas component into a BH
541: in the more massive galaxies, while a star cluster
542: was formed in galaxies with a shallower gravitational
543: potential, where the central accumulation of gas
544: occurred more slowly.
545:
546: However, many questions remain open. Some relatively
547: massive ellipticals do have NSCs (Rossa \etal\ 2006;
548: although they have not been found in galaxies brighter
549: than $M_B \sim -20$~mag; Ferrarese \etal\ 2006).
550: The range in which NSCs and SMBHs overlap may span
551: several orders of magnitude: a SMBH with $M_{\rm BH} \sim 3\times 10^6$~\msun\
552: is the CMO in the center of the Milky Way, and a SMBH
553: with a mass $\sim 10^5$~\msun\ is present in NGC~4395, but
554: NSCs as massive as a few $10^8$~\msun\ have been
555: found in some elliptical galaxies (Rossa \etal\ 2006).
556: It is not known whether the same galaxy can harbor
557: both an SMBH and a NSC, or if they are mutually
558: exclusive. It is also not yet clear which of
559: the two classes of objects defines the more fundamental
560: correlation with the galactic bulge; it was suggested
561: (Rossa \etal\ 2006) that NSCs in large galaxies
562: may be a factor of 3 more massive than SMBHs,
563: for a given bulge mass. Furthermore, it is not known
564: whether a linear relation between bulge and CMO
565: masses may extend all the way down to bulgeless
566: Scd galaxies (implying the existence of CMOs
567: with masses as low as $10^3-10^5$~\msun), or if
568: the relation breaks down for CMO masses \LA $10^6$~\msun.
569:
570: From its optical brightness and colors, we estimate
571: a stellar mass of $\sim$2$\times$10$^6$~\msun\ for the NSC in \ngc,
572: comparable to the mass of the SMBH in the Milky Way,
573: even though \ngc\ is bulgeless. Although our
574: X-ray study suggests that the nuclear X-ray source
575: is not from a SMBH, this does not rule out the presence
576: of a SMBH inside the NSC: it may simply be quiescent.
577: In that case, it would be among the faintest SMBHs,
578: with a luminosity \LA 10$^{-10}$~$L_{\rm Edd}$.
579: In a work currently in preparation, we shall
580: use an X-ray survey of galactic nuclei to determine
581: whether galaxies with NSCs may also have signs
582: of AGN activity.
583:
584: Even if we accept that a late-type galaxy such as \ngc\
585: formed a NSC but no SMBH, there is a strong possibility
586: that the NSC contains stellar-mass BHs formed via
587: ordinary stellar processes, or perhaps an intermediate-mass BH (IMBH)
588: formed via runaway core collapse and stellar coalescence
589: (if the NSC was sufficiently compact). Then, those BHs
590: could have grown over a Hubble time, via accretion
591: during large-scale gas inflows towards the galactic center,
592: via coalescence of stellar mass BHs inside the NSC,
593: or via orbital decay and coalescence of accreted
594: satellite galaxies or primordial halo remnants.
595: Processes involving BH mergers may lead to the displacement
596: or escape of the merged BH due to gravitational radiation
597: recoil, and to the formation of a lower-density stellar core
598: (Merritt \etal\ 2004).
599: It is possible that some late-type galaxies have
600: a three-tier hierarchical relation between a bulge,
601: a NSC, and an IMBH inside the NSC. Theoretical arguments
602: have suggested that if an IMBH forms via core collapse
603: in the core of a massive star cluster, its mass
604: should be $\sim$0.2\% of the host cluster mass, coincidentally
605: similar to the mass ratio between the bulge and its CMO.
606: %(Freitag et al .... Portegies Zwart et al...)
607:
608: In principle, the effective radius of a NSC may provide
609: clues on its dynamical processes. In \ngc, the NSC
610: is older and less compact ($r_e = 11.7$~pc)
611: than most other NSCs in late-type spirals (typical
612: $r_e \sim 3.5$~pc). If the cluster was formed with
613: such a large $r_e$, it would not have been compact
614: enough to form an IMBH via core collapse (Portegies~Zwart
615: \etal\ 2004). Alternatively, one may speculate
616: that the NSC was originally more compact and
617: has evolved to such a large radius because
618: of the presence of one or more BHs in its core
619: (Ebisuzaki \etal\ 1991; Milosavljevi\'{c} \etal\ 2002;
620: Graham 2004).
621: Other physical processes not related to BH coalescence
622: may also cause a star cluster to expand. For example,
623: studies of clusters in the LMC show clusters
624: are born compact and evolve to a range of \re\ as they age
625: (Elson \etal\ 1989; Mackey \& Gilmore 2003).
626: But studies also show
627: this evolution is not due to differences in IMF (deGrijs \etal\ 2002),
628: time-varying external tidal fields, nor differences in the
629: number of hard primordial binaries present (Wilkinson \etal\ 2003).
630: If no massive BH formed, the cluster could still evolve to large \re.
631: For the LMC clusters (none of which are thought to host a massive BH!),
632: \re\ evolution
633: is likely due to differences in star formation efficiencies with
634: lower efficiencies tending to expel gas before a dense stellar core
635: (with a population of massive stars) can form
636: and leading eventually to cluster expansion
637: (Goodwin 1997; Vine \& Bonnell 2003).
638: If evolution is through this mechanism, then we would expect a
639: smooth distribution of cluster \re\ values that included the value
640: $r_e=11.7$~pc found here for \ngc.
641: This does not appear to be the case (cf. Fig.~4 of B\"{o}ker \etal\ 2004)
642: although the B\"{o}ker \etal\ sample is not large.
643:
644:
645: Another difference between NSCs with or without
646: a relatively massive BH is the role of feedback.
647: More powerful feedback is expected from a nuclear
648: IMBH or SMBH, leading to gas expulsions, perhaps
649: large-scale outflows, and cyclic episodes of star
650: formations. In this sense, the distribution
651: of stellar populations in a NSC provides a fossil
652: record of past galactic activity, and may be used
653: to reconstruct episodes of galactic mergers
654: or phases of nuclear activity.
655: However, in the case
656: of a relatively low-mass NSC such as the one
657: in \ngc, we find that normal stellar winds and SNe
658: can provide enough energy to expel cool interstellar
659: gas from the cluster and quench star formation,
660: without the need of AGN feedback.
661:
662: Interestingly, \ngc\ is one of those rare galaxies with gas column densities
663: below the critical value needed to sustain activity but has a normal
664: star formation rate (Martin \& Kennicutt 2001).
665: Perhaps a strong outflow from the central region has helped trigger
666: star formation at larger radii.
667: Outflow may also account for some of the anomalous \hi\ discovered in the
668: halo of \ngc\ (Fraternalli \etal\ (2002, 2004).
669: These points will be addressed in a future paper (Yukita \etal, in preparation).
670:
671: The time interval since the last star formation episode is longer
672: for the NSC in \ngc\ than is typical for clusters in late-type spirals.
673: %
674: Davidge \& Courteau (2002) also noted the lack of star formation activity
675: more recent than $\sim$100~Myr in the inner disk of \ngc\ and suggested
676: it may be due to expulsion of potential star-forming material from
677: the central regions.
678: Indeed, active star formation is ongoing in several giant \hii\ regions
679: (Drissen \etal\ 1999; Yukita \etal, in preparation) surrounding the
680: nucleus of \ngc\
681: but the NSC itself contains perhaps the oldest collection of stars in
682: the region.
683:
684: \begin{thebibliography}{}
685:
686: \bibitem[]{}
687: Beswick, R.J., \etal\ 2005, ApJ, 623, L21
688: \bibitem[]{}
689: B\"{o}ker, T., \etal\ 2002, AJ, 123, 1389
690: \bibitem[]{}
691: B\"{o}ker, T., \etal\ 2004, AJ, 127, 105
692: \bibitem[]{}
693: Buccheri, R. \etal\ 1983, A\&A, 128, 245
694: \bibitem[]{}
695: Carollo, C. M., Stiavelli, M., \& Mack, J. 1998, AJ, 116, 68
696: \bibitem[]{}
697: Davidge, T. J. \& Courteau, S. 2002, AJ, 123, 1438
698: \bibitem[]{}
699: deGrijs, R., \etal\ 2002, MNRAS 337, 597
700: \bibitem[]{}
701: Drissen, L., Roy, J. -R., Moffat, A. F. J., \& Shara, M. M. 1999, AJ, 117, 1249
702: \bibitem[]{}
703: Dubus, G., \& Rutledge, R. E. 2002, MNRAS, 336, 901
704: \bibitem[]{}
705: Elson, L. A. W., Freeman, K. C., \& Lauer, T. R. 1989, ApJ, 347, L69
706: \bibitem[]{}
707: Ebisuzaki, T., Makino, J., \& Okumura, S. K. 1991, Nature, 354, 212
708: \bibitem[]{}
709: Ferrarese, L., \etal\ 2006, ApJ, 644, L21
710: \bibitem[]{}
711: Ferrarese, L. \& Merritt, D. 2000, ApJ, 539, L9
712: \bibitem[]{}
713: Filippenko, A. V., \& Sargent, W. L. W. 1989, ApJ, 342, L11
714: \bibitem[]{}
715: Filippenko, A. V., \& Ho, L. C. 2003, ApJ, 588, L13
716: \bibitem[]{}
717: Foschini, L. \etal\ 2004, A\&A, 416, 529
718: \bibitem[]{1388}
719: Fraternali, F., van~Moorsel, G., Sancisi, R., \& Oosterloo, T. 2002, AJ, 123, 3124
720: \bibitem[]{1388}
721: Fraternali, F., Oosterloo, T., \& Sancisi, R. 2004, A\&A, 424, 485
722: \bibitem[]{}
723: Freeman, K. C. 1993, in ASP Conf Ser 48, The Globular Cluster-Galaxy Connection, ed. G. H. Smith \& J. P. Brodie (San Francisco:ASP), 608
724: \bibitem[]{}
725: Gebhardt, K., \etal\ 2000, ApJ, 539, L13
726: \bibitem[]{}
727: Gebhardt, K., \etal\ 2001, AJ, 122, 2469
728: \bibitem[]{}
729: Gordon, K. D., Hanson, M. M., Clayton, G. C., Rieke, G. H., \& Misselt, K. A. 1999, ApJ, 519, 165
730: \bibitem[]{}
731: Goodwin, S. P. 1997, MNRAS 286, 669
732: \bibitem[]{}
733: Graham, A. W., Erwin, P., Caon, N., \& Trujillo, I. 2001, ApJ, 563, L11
734: \bibitem[]{}
735: Graham, A. W. 2004, ApJ, 613, L33
736: \bibitem[]{}
737: Greiner, J., Cuby, J. G., \& McCaughrean, M. J. 2001, Nature, 414, 522
738: \bibitem[]{}
739: H\"{a}ring, N. \& Rix, H. -W. 2004, ApJ, 604, L89
740: \bibitem[]{}
741: Iwasawa, K., Fabian, A. C., Almaini, O., Lira, P., Lawrence, A., Hayashida, K., Inoue, H. 2000, MNRAS, 318, 879
742: \bibitem[]{}
743: King, A. R. \& Dehnen, W. 2005, MNRAS, 357, 275
744: \bibitem[]{}
745: Kormendy, J. \& McClure, R. D. 1993, AJ, 105, 1793
746: \bibitem[]{}
747: Kormendy, J. \& Richstone, D. 1995, ARA\&A, 33, 581
748: \bibitem[]{}
749: Krist, J. E., \& Hook, R.N. 1997, in
750: The 1997 HST Calibration Workshop with a New Generation of Instruments, p. 192
751: \bibitem[]{}
752: La~Parola, V., Damiani, F., Fabbiano, G., \& Peres, G. 2003, ApJ, 583, 758
753: \bibitem[]{}
754: Larsen, S. S. 1999, A\&AS, 139, 393
755: \bibitem[]{}
756: Layden, A. C. \& Sarajendini, A. 2000, AJ, 119, 1760
757: \bibitem[]{}
758: Leitherer, C. \etal\ 1999, ApJS, 123, 3
759: \bibitem[]{}
760: Long, K. S., Charles, P. A., \& Dubus, G. 2002, ApJ, 569, 204
761: \bibitem[]{}
762: Mackey, A. D. \& Gilmore, G. F. 2003, MNRAS, 338, 85
763: \bibitem[]{}
764: Madore, B. F. \& Freedman, W. L. 1991, PASP, 103, 933
765: \bibitem[]{}
766: Magorrian, J., \etal\ 1998, AJ, 115, 2285
767: \bibitem[]{}
768: Makishima, K., \etal\ 1986, ApJ, 308, 635
769: \bibitem[]{}
770: Makishima, K., \etal\ 2000, ApJ, 535, 632
771: \bibitem[]{}
772: Martin, C. L., \& Kennicutt, R. C. 2001, ApJ, 555, 301
773: \bibitem[]{}
774: Merritt, D., Milosavljevi\'{c}, M., Favata, M., Hughes, S. A., \& Holz, D. E. 2004, ApJ, 607, L9
775: \bibitem[]{}
776: Moran, E. C., \etal\ 1999, PASP, 111,801
777: \bibitem[]{}
778: Moran, E. C., Eracleous, M., Leighly, K. M., Chartas, G., Filippenko, A. V.,
779: Ho, L. C., Blanco, P. R. 2005, AJ, 129, 2108
780: \bibitem[]{}
781: Milosavljevi\'{c}, M., Merritt, D., Rest, A., \& van den Bosch, F. C. 2002,
782: MNRAS, 331, L51
783: \bibitem[]{}
784: Milosavljevi\'{c}, M. 2004, ApJ, 605, L13
785: \bibitem[]{}
786: O'Connell, R. W. 1983, ApJ, 267, 80
787: \bibitem[]{}
788: Page, M. J., Davis, S. W., \& Salvi, N. J. 2003, MNRAS, 343, 1241
789: \bibitem[]{}
790: Portegies Zwart, S. F., Dewi, J., \& Maccarone, T. 2004, MNRAS, 355, 413
791: \bibitem[]{}
792: Remillard, R. A., \& McClintock, J. E. 2006, ARA\&A, 44, 49
793: \bibitem[]{}
794: Rossa, J., \etal\ 2006, AJ, 132, 1074
795: \bibitem[]{}
796: Schlegel, D. J., Finkbeiner, D. P., \& Davis, M. 1998, ApJ, 500, 525
797: \bibitem[]{}
798: Shih, D. C., Iwasawa, K., \& Fabian, A. C. 2003, MNRAS, 341, 973
799: \bibitem[]{}
800: Sirianni, M., \etal\ 2005, PASP, 117, 1049
801: \bibitem[]{}
802: Swartz, D. A., Ghosh, K. K., Tennant, A. F., \& Wu, K. 2004, ApJS, 154, 519
803: \bibitem[]{}
804: Tremaine, S. \etal\ 2002, ApJ, 574, 740
805: \bibitem[]{}
806: Trinchieri, G., Fabbiano, G., \& Peres, G. 1988, 325,531
807: \bibitem[]{}
808: van~Paradijs, J. \& McClintock, J. E. 1995, in X-ray Binaries, eds. W. H. G. Lewin, J. van Paradijs, and E. P. J. van den Heuvel (Cambridge: Cambridge Univ. Press), p. 58
809: \bibitem[]{}
810: Vine, S. G., \& Bonnell, I. A. 2003, MNRAS 342, 314
811: \bibitem[]{}
812: Walcher, C. J., \etal\ 2006, ApJ, 618, 237
813: \bibitem[]{}
814: Wehner, E. H., \& Harris, W. E. 2006, ApJ, 644, L17
815: \bibitem[]{}
816: Wilkinson, M. I., Hurley, J. R., Mackey, A. D., Gilmore, G. F., \& Tout, C. A.
817: 2003, MNRAS, 343, 1025
818:
819: \end{thebibliography}
820:
821: \begin{center}
822: \begin{tabular}{crlcr}
823: \multicolumn{5}{c}{{\sc TABLE 1}} \\
824: \multicolumn{5}{c}{{\sc X-ray Observations of NGC 2403 and its Nuclear Source Luminosity}} \\
825: %
826: \hline \hline
827: No. & \multicolumn{1}{c}{Mission/Instrument} & \multicolumn{1}{c}{Date/Identifier} & \multicolumn{1}{c}{$L_{\rm X}$(0.5-8.0 keV)} & \multicolumn{1}{c}{$t_{exp}$} \\
828: & & & (10$^{37}$ \ergl ) & (ks) \\
829: \hline
830: 1 & \cxo/ACIS-S & 2001-04-17/2014 & $<0.049$ & 36.1 \\
831: %%%\cxo/ACIS-S & 2004-08-09/4627 & --- & 45.1 \\
832: 2 & \xmm/MOS$+$PN & 2003-04-30/0150651101 & $<2.8$ & 8.0 \\
833: 3 & \xmm/MOS$+$PN & 2003-09-11/0150651201 & $69.8^{+11.3}_{-8.2}$ & 7.5 \\
834: 4 & \cxo/ACIS-S & 2004-08-23/4628 & $<0.005$ & 47.1 \\
835: 5 & \xmm/MOS$+$PN & 2004-09-12/0164560901 & $<1.0$ & 60.0 \\
836: 6 & \cxo/ACIS-S & 2004-10-03/4629 & $2.9^{+0.1}_{-2.5}$ & 45.1 \\
837: 7 & \cxo/ACIS-S & 2004-12-22/4630 & $18.7^{+0.9}_{-7.0}$ & 50.6 \\
838: \hline
839: %\multicolumn{4}{l}{}
840: \end{tabular}
841: %}
842: \end{center}
843:
844: \begin{center}
845: \begin{tabular}{cccc}
846: \multicolumn{4}{c}{{\sc TABLE 2}} \\
847: \multicolumn{4}{c}{{\sc X-Ray Spectral Fit Parameters}} \\
848: %
849: \hline \hline
850: & Observation 3 & Observation 6 & Observation 7 \\
851: \hline
852: Fitted energy range (keV) & 0.3 - 10.0 & 0.5 - 2.0 & 0.5 - 3.5 \\
853: \hline
854: \multicolumn{4}{c}{Disk Blackbody Model} \\
855: $n_H$ (10$^{21}$ cm$^{-2}$)$^a$ & $0.9^{+0.2}_{-0.2}$ & $0.4^{+1.0}_{-0.0}$ & $0.4^{+0.7}_{-0.0}$ \\
856: $T_{\rm in}$ (keV) & $1.11^{+0.09}_{-0.09}$ & $0.46^{+0.06}_{-0.12}$ & $0.81^{+0.10}_{-0.13}$\\
857: $R_{\rm in}$ (km)$^b$ & $47.0^{+7.9}_{-6.6}$ & $60.1^{+75.5}_{-15.1}$ & $45.7^{+20.4}_{-9.1}$\\
858: $M$/\msun\ & 5.4 & 7.2 & 5.4 \\
859: $\dot{M}$ (10$^{-7}$ \msun\ yr$^{-1}$) & 1.24 & 0.05 & 0.30 \\
860: $L_{\rm disk, bol}$ (10$^{37}$ \ergl ) & 85.4 & 4.52 & 23.7 \\
861: $\chi^2$/DOF$^c$ & 73.8/97 & 115.6/101 & 215.7/204 \\
862: \hline
863: \multicolumn{4}{c}{Power Law Model} \\
864: $n_H$ (10$^{21}$ cm$^{-2}$) & $3.0^{+0.4}_{-0.4}$ & $2.2^{+2.1}_{-1.8}$ & $2.1^{+1.1}_{-1.0}$ \\
865: $\Gamma$ & $2.22^{+0.13}_{-0.13}$ & $3.05^{+1.24}_{-1.15}$ & $2.28^{+0.42}_{-0.40}$\\
866: $\chi^2$/DOF$^c$ & 91.7/97 & 116.5/101 & 218.2/204 \\
867: \hline
868: \multicolumn{4}{l}{$^a$Fixed at Galactic value for Observations 6 \& 7}\\
869: \multicolumn{4}{l}{$^b$Assumes $\cos \theta \sim 1$}\\
870: \multicolumn{4}{l}{$^c$C-statistic/Number of data bins for Observations 6 \& 7}
871: \end{tabular}
872: %}
873: \end{center}
874:
875:
876: \end{document}
877:
878: