1: \documentclass[12pt,preprint]{aastex}
2:
3: \newcommand{\mdot}{\mbox{$\dot M$}{}}
4:
5: \begin{document}
6: \title{Diffuse X-rays from the Arches and Quintuplet Clusters}
7:
8: \author{Gabriel Rockefeller\altaffilmark{1,2}, Christopher L.
9: Fryer\altaffilmark{1,2}, Fulvio Melia\altaffilmark{1,3,4}, and Q.
10: Daniel Wang\altaffilmark{5}}
11:
12: \altaffiltext{1}{Department of Physics, The University of Arizona,
13: 1118 E 4th St, Tucson, AZ 85721}
14: \altaffiltext{2}{Theoretical Division, Los Alamos National Laboratory,
15: P.O. Box 1663, Los Alamos, NM 87545}
16: \altaffiltext{3}{Steward Observatory, The University of Arizona, 933 N
17: Cherry Ave, Tucson, AZ 85721}
18: \altaffiltext{4}{Sir Thomas Lyle Fellow and Miegunyah Fellow.}
19: \altaffiltext{5}{Department of Astronomy, University of Massachusetts,
20: 710 N Pleasant St, Amherst, MA 01003}
21:
22: \begin{abstract}
23: The origin and initial mass function of young stellar clusters near
24: the Galactic center are still poorly understood. Two of the more
25: prominent ones, the Arches and Quintuplet clusters, may have formed
26: from a shock-induced burst of star formation, given their similar
27: age and proximity to each other. Their unusual mass distribution,
28: however, may be evidence of a contributing role played by other
29: factors, such as stellar capture from regions outside the clusters
30: themselves. Diffuse X-ray emission from these sources provides us
31: with a valuable, albeit indirect, measure of the stellar mass-loss
32: rate from their constituents. Using recent data acquired with
33: \textit{Chandra}, we can study the nature and properties of the
34: outflow to not only probe the pertinent physical conditions, such as
35: high metallicity, the magnetic field, and so forth, but also to
36: better constrain the stellar distribution within the clusters, in
37: order to identify their formative history. In this paper, we
38: present a set of three-dimensional smoothed particle hydrodynamics
39: simulations of the wind-wind interactions in both the Arches and
40: Quintuplet clusters. We are guided primarily by the currently known
41: properties of the constituent stars, though we vary the mass-loss
42: rates in order to ascertain the dependence of the measured X-ray
43: flux on the assumed stellar characteristics. Our results are
44: compared with the latest observations of the Arches cluster. Our
45: analysis of the Quintuplet cluster may be used as a basis for
46: comparison with future X-ray observations of this source.
47: \end{abstract}
48:
49: \keywords{galaxies: clusters: individual (Arches,
50: Quintuplet)---Galaxy: center---radiation mechanisms: thermal---shock
51: waves---stars: winds---X-rays: diffuse}
52:
53: \section{Introduction}
54:
55: Understanding the environment's role in star formation and, in turn,
56: the feedback exerted by star formation on the Galactic environment, is
57: a problem of significance to several fields in astronomy, from the
58: creation of compact objects to the formation of galaxies. The
59: Galactic center, with its relatively high magnetic field strength,
60: clouds with high particle density, and large velocity dispersions,
61: provides an ideal environment to study star formation under extreme
62: conditions \citep{M93,MF01}. This type of investigation can benefit
63: from the existence of several stellar clusters in this region,
64: including the Arches and Quintuplet clusters. Learning more about
65: their stellar constituents, and possibly their formative history, may
66: even give us a glimpse into the emergence of objects that will
67: ultimately populate the central parsec of the Galaxy
68: \citep[e.g.,][]{G01,M03}.
69:
70: The Arches and Quintuplet clusters have been studied over a range of
71: wavelengths, from radio to X-ray. X-rays provide a unique window for
72: investigating both the formation of binaries (point sources from
73: binary interactions) and the wind interactions within the entire
74: cluster (diffuse emission). As \citet{Rock04} have shown, the diffuse
75: X-ray emission is a sensitive measure of the mass-loss rate of stars
76: in mutually interactive situations. Stellar mass-loss remains one of
77: the largest uncertainties in stellar evolution \citep[e.g.,][]{WL99}.
78: X-ray observations of these clusters represent a unique probative tool
79: for studying the winds produced by high-metallicity systems.
80:
81: In this paper, we model the propagation and interaction of winds from
82: stars in both the Arches and Quintuplet clusters, calculating the
83: X-ray fluxes arising from the consequent shocked gas. \citet{Y02}
84: serendipitously discovered the Arches cluster with \textit{Chandra}
85: and identified five components of X-ray emission, which they labeled
86: A1--A5, though only A1--A3 seem to be directly associated with the
87: cluster \citep[see Figure 2,][]{Y02}. A1 is apparently associated
88: with the core of the cluster, while A2 is located $\sim
89: 10^{\prime\prime}$ northwest of A1. A1 and A2 are partially resolved,
90: while A3 is a diffuse component that extends beyond the boundary of
91: the cluster; \citet{Y02} speculated that some or all of A1--A3 may be
92: produced by interactions of winds from stars in the system. Analysis
93: of additional \textit{Chandra} observations that covered the Arches
94: cluster and first results from the Quintuplet cluster are presented by
95: \citet{W03} and \citet{L04}; these new observations resolve A1 into
96: two distinct components, labeled A1N and A1S, and indicate that A1N,
97: A1S, and A2 are all point-like X-ray sources.
98:
99: Several efforts have already been made to study the X-ray
100: emission from clusters. \citet{OGU97} and \citet{C00} performed
101: analytic calculations to estimate the diffuse emission from these
102: systems. The interaction of winds in the Arches cluster has been
103: simulated by \citet{R01} using the \mbox{``yguaz\'{u}-a''} adaptive
104: grid code described in \citet{R00}. However, all previous work
105: focused exclusively on the Arches cluster, and even the detailed
106: simulations assumed identical large values for the mass-loss rates
107: ($\dot M = 10^{-4} M_\odot$~yr$^{-1}$) and wind velocities ($v_{wind}
108: = 1,500$~km~s$^{-1}$) of the constituent stars. In this paper, we
109: present the results of simulations of both the Arches and Quintuplet
110: clusters, using detailed radio flux measurements (where available) and
111: spectral classifications to pin down the expected mass-loss rates of
112: stars in each system. We then compare our results to the most recent
113: X-ray observations of these two clusters, including new data presented
114: in this paper.
115:
116: A summary of relevant properties of each cluster is presented below.
117: We describe our numerical technique, including the characteristics of
118: the wind sources and the gravitational potential of the clusters, in
119: \S\ \ref{sect.physicalsetup} of the paper. The new observations are
120: described in \S\ \ref{sect.obs}. A comparison of the theoretical
121: results with the data is made in \S\ \ref{sect.results}, and the
122: relevance to the Galactic center conditions is discussed in \S\
123: \ref{sect.discussion}.
124:
125: \subsection{The Arches Cluster}\label{sect.Arches}
126:
127: The Arches stellar cluster is an exceptionally dense aggregate of
128: stars located at $l=0.12^\circ$, $b=0.02^\circ$, about $11^\prime$ in
129: projection from the Galactic center \citep[see e.g.][]{N95,C96,F02}.
130: The cluster is apparently a site of recent massive star formation; it
131: contains numerous young emission-line stars which show evidence of
132: strong stellar winds. Using near-IR color-magnitudes and K band
133: counts, \citet{S98} estimated that at least 100 cluster members are O
134: stars with masses greater than $20\;M_\odot$ and calculated a total
135: cluster mass of $\sim(1.5$--$6)\times10^4\;M_\odot$. \citet{F99b}
136: used \textit{HST} NICMOS observations to determine the slope of the
137: initial mass function (IMF) of the Arches cluster and calculated a
138: cluster mass of $\sim 10^4\;M_\odot$, with possibly 160 O stars and an
139: average mass density of $\sim 3\times 10^5\;M_\odot$~pc$^{-3}$.
140:
141: The 14 brightest stars of this cluster have been identified with JHK
142: photometry and Br$\alpha$ and Br$\gamma$ hydrogen recombination lines,
143: showing that they have the characteristic colors and emission lines of
144: Of-type or Wolf-Rayet (WR) and He~\textsc{i} emission-line stars.
145: \citet{N95} inferred from the strength of the Br$\alpha$ and
146: Br$\gamma$ line fluxes that these stars are losing mass at a
147: prodigious rate, $\dot M\sim 2\times 10^{-5}\;M_\odot$~yr$^{-1}$, in
148: winds moving at $\sim 10^3$~km~s$^{-1}$. \citet{C96} confirmed the
149: presence of young, massive stars using $K$-band spectroscopy; they
150: identified 12 stars in the cluster with spectra consistent with
151: late-type WN/Of objects, with mass-loss rates $\dot M\sim
152: (1$--$20)\times 10^{-5}\;M_\odot$~yr$^{-1}$ and wind velocities
153: $v_\infty\sim 800$--$1,200$~km~s$^{-1}$.
154:
155: Follow-up Very Large Array (VLA) observations at centimeter
156: wavelengths of the brightest stars in the cluster have solidified the
157: detection of powerful ionized stellar winds. Using the observed
158: 8.5~GHz flux densities from 8 sources in the Arches cluster and the
159: relationship between flux density and mass-loss rate derived by
160: \citet{P75},
161: \begin{equation}\label{math-mdot}
162: \dot M=(5.9\times 10^{-5}\;M_\odot\,{\hbox{yr}}^{-1})\left({S_{8.5}\over
163: 1\,\hbox{mJy}}\right)^{3/4}\left({v_\infty\over 500\;\hbox{km}\;{\hbox{s}}^{-1}}
164: \right)\left({d\over 8\;\hbox{kpc}}\right)^{3/2}\;,
165: \end{equation}
166: where $S_{8.5}$ is the 8.5~GHz flux density, $v_\infty$ is the wind
167: terminal velocity, and $d$ is the distance to the source ($\sim
168: 8$~kpc, for stars in the Arches cluster), \citet{L01} calculated mass
169: loss rates $\dot M=(3$--$17)\times 10^{-5}\;M_\odot$~yr$^{-1}$,
170: assuming a wind electron temperature $T\sim 10^4$~K, $Z = 1$, and a
171: mean molecular weight $\mu = 2$. The Wolf-Rayet phase is short-lived,
172: but while in this mode, stars dominate the mass ejection within the
173: cluster.
174:
175: In an attempt to represent both the identified stellar wind sources
176: and the population of stars likely to be producing significant but
177: currently undetected winds, we include 42 wind sources (listed in
178: Table~\ref{Arches.srcs} and shown in Figure~\ref{fig-arches}) in our
179: simulations of the Arches cluster. The stars labeled AR1--AR9
180: correspond to the 9 radio sources identified by \citet{L02}. The
181: first 29 stars (labeled 1--29) in Table~\ref{Arches.srcs} have mass
182: estimates greater than $60\;M_\odot$ \citep{F02}, and are likely the
183: most powerful sources of wind in the cluster. The remaining 13 stars
184: used in the simulations have masses less than $60\;M_\odot$ but
185: greater than $25\;M_\odot$ and are located on the north side of the
186: cluster; they are included to better represent the spatial
187: pattern of X-ray emission around the core of the cluster. Based on
188: the broadening of the Br$\gamma$ line observed by \citet{C96}, stars
189: in the Arches cluster simulations are assigned wind velocities of
190: $1,000$~km~s$^{-1}$. The stars labeled AR1--AR9, which have observed
191: 8.5~GHz flux densities, are assigned mass-loss rates according to
192: Equation~\ref{math-mdot}. Stars with no associated 8.5~GHz detection
193: but with masses larger than $60\;M_\odot$ are assigned a mass-loss
194: rate of $3\times 10^{-5}\;M_\odot$~yr$^{-1}$, which is equal to the
195: lowest mass-loss rate inferred from the weakest observed 8.5~GHz
196: signal from the Arches cluster \citep{L02}. Stars with masses less
197: than $60\;M_\odot$ are assigned a mass-loss rate of $3\times
198: 10^{-6}\;M_\odot$~yr$^{-1}$; their winds will have little effect on
199: the overall luminosity but may alter the shape of the X-ray-emitting
200: region.
201:
202: \subsection{The Quintuplet Cluster}\label{sect.Quintuplet}
203:
204: Slightly further north of Sgr A*, the Quintuplet cluster is located at
205: $l=0.16^\circ$, $b=0.06^\circ$. The cluster has a total estimated
206: mass of $\sim 10^4\;M_\odot$ and a mass density of $\sim
207: 10^3\;M_\odot$~pc$^{-3}$ \citep{F99a}. Like the objects in the Arches
208: cluster, the known massive stars in the Quintuplet cluster have
209: near-IR emission-line spectra indicating that they too have evolved
210: away from the zero-age main sequence and now produce high-velocity
211: stellar winds with terminal speeds of $500$--$1,000$~km~s$^{-1}$.
212:
213: \citet{F99a} obtained K-band spectra of 37 massive stars in the
214: Quintuplet cluster and found that 33 could be classified as WR stars,
215: OB supergiants, or luminous blue variables (LBVs), implying a range of
216: wind mass-loss rates $\dot M\sim (0.1$--$6.6)\times
217: 10^{-5}\;M_\odot$~yr$^{-1}$. VLA continuum images at 6~cm and 3.6~cm
218: of the Sickle and Pistol H~\textsc{ii} regions reveal eight point
219: sources located in the vicinity of the Quintuplet cluster, including
220: the radio source at the position of the Pistol nebula \citep{L99}.
221: These are labeled QR1 through 7, and the Pistol star, in
222: Figure~\ref{fig-quint} below.
223:
224: The near-IR counterparts of QR4 and QR5 are hot, massive stars with
225: high mass-loss rates, one an OB~\textsc{i} supergiant and the other a
226: WN9/Ofpe \citep{F99a}. The sources QR1, QR2, and QR3 also have
227: spectral indices consistent with stellar wind sources, but they have
228: no obvious NICMOS stellar counterparts. \citet{L99} speculated that
229: this may be due to variable extinction associated with a dense
230: molecular cloud located in front of the cluster. Given the
231: uncertainty, and possible variation, in stellar identification for
232: these 5 objects, we therefore adopt a value of $\sim 500$~km~s$^{-1}$
233: (typical in OB supergiants) for the speed of their wind. The Pistol
234: star, on the other hand, is a prominent source in the near-IR NICMOS
235: image, and is evidently a luminous blue variable \citep{F98}
236: possessing a powerful stellar wind, though with a terminal speed of
237: only $v_\infty\sim 100$~km~s$^{-1}$.
238:
239: In our simulations of the Quintuplet cluster, we include 31 massive
240: stars with spectral classifications identified by \citet{F99a}, using
241: estimates for the wind parameters of each star according to its
242: spectral type. Wind velocities are determined according to the broad
243: spectral type of each star: we assume a velocity of
244: $1,000$~km~s$^{-1}$ for winds from WR stars, $500$~km~s$^{-1}$ winds
245: for OB supergiants, and $100$~km~s$^{-1}$ for the Pistol star, a LBV.
246:
247: For those stars that are radio sources \citep[QR1--QR3, QR6, and QR7
248: from][]{L99}, mass-loss rates are determined according to
249: Equation~\ref{math-mdot}. For the rest of the stars, we estimate the
250: mass-loss based on the spectral classifications by \citet{F99a}. For
251: OB stars, the mass-loss rate was assumed to be
252: $10^{-5}\;M_\odot$~yr$^{-1}$ for stars with classification higher than
253: BO. For smaller stars, the mass-loss rate was assumed to be
254: $10^{-6}\;M_\odot$~yr$^{-1}$. For Wolf-Rayet stars, we use a
255: luminosity-mediated mass-loss relation \citep{WL99}:
256: \begin{equation}
257: log \left( -\dot{M}_{\rm WR}\over M_\odot\;yr^{-1} \right) =
258: k+1.5 log \left(L\over L_\odot\right) - 2.85 X_s,
259: \end{equation}
260: where $L$ is the stellar luminosity from \citet{F99a}, $X_s$ is the
261: hydrogen mass fraction, and $k$ is a constant which we calibrated
262: using our radio-determined mass-loss rates. The locations and wind
263: parameters of the stellar wind sources are summarized in
264: Table~\ref{Quint.srcs}, and their relative positions are shown in
265: Figure~\ref{fig-quint}.
266:
267: \section{The Physical Setup}\label{sect.physicalsetup}
268:
269: Our calculations use the three-dimensional smoothed particle
270: hydrodynamics (SPH) code discussed in \citet{FW02} and \citet{War04},
271: modified as described in \citet{Rock04} to include stellar wind
272: sources. The gridless Lagrangean nature of SPH allows us to
273: concentrate spatial resolution near shocks and model gas dynamics and
274: wind-wind interactions on length scales that vary by several orders of
275: magnitude within a single calculation.
276:
277: We assume that the gas behaves as an ideal gas, according to a
278: gamma-law ($\gamma = 5/3$) equation of state. The effect of
279: self-gravity on the dynamics of the gas should be negligible compared
280: to the effect of the central cluster potential; we calculate
281: gravitational effects by approximating the potential of each cluster
282: with a Plummer model.
283:
284: The computational domain for the simulation of each cluster is a cube
285: approximately $6$~pc on a side, centered on the middle of the cluster.
286: To simulate ``flow-out'' conditions, particles passing through the
287: outer boundary are removed from the simulation. The initial
288: conditions assume that the space around and within each cluster is
289: empty; massive stars in each cluster then inject matter into the
290: volume of solution via winds as the calculation progresses. The
291: number of particles in each simulation initially grows rapidly, but
292: reaches a steady number ($\sim 6.6$ million particles for simulations
293: of the Arches cluster and $\sim 3.3$ million particles for the
294: Quintuplet simulations, since there are fewer identified wind sources
295: in the latter) when the addition of particles from wind sources is
296: compensated by the loss of particles through the outer boundary of the
297: computational domain.
298:
299: \subsection{Cluster Potential}
300:
301: We model the gravitational potential $\Phi$ of a cluster using a
302: Plummer model \citep{P11},
303: \begin{equation}
304: \Phi(r) = \frac{-GM}{\sqrt{r^{2}+b^{2}}}\;,
305: \end{equation}
306: where $M$ is the total mass of the cluster. The radial density
307: profile assumed for the cluster is therefore
308: \begin{equation}
309: \rho(r)=\rho_{0}{\left(1+\frac{r^{2}}{b^{2}}\right)}^{-5/2}\;,
310: \end{equation}
311: where the central density $\rho_{0}$ is
312: \begin{equation}
313: \rho_{0}=\frac{3M}{4\pi b^{3}}\;.
314: \end{equation}
315: We note that the spatial distributions of massive stars in the
316: clusters are not entirely consistent with the distribution implied by
317: the Plummer potential; for example, the average projected distance of
318: both the set of massive stars and the set of all observed stars in the
319: Arches cluster is roughly twice the average distance predicted by a
320: Plummer model based on the estimated total mass and central density of
321: the cluster. We include the Plummer potential to approximate the
322: combined gravitational influence of the entire cluster, including the
323: estimated mass of stars too small or dim to be observed.
324:
325: \citet{F99a} provide estimates of the total mass of each cluster by
326: measuring the mass of observed stars, assuming a Salpeter IMF slope,
327: and extrapolating down to $1\;M_\odot$---observed stars account for at
328: most 25\% of the mass of the Arches cluster and 16\% of the mass of
329: the Quintuplet cluster. They also estimate the density of each
330: cluster by determining the volume of the cluster from the average
331: projected distance of stars from the cluster center, and dividing the
332: total mass by this volume; because the values are calculated using the
333: total cluster mass but only the average projected cluster radius, the
334: density estimates are probably closer to the central densities than
335: the average densities. We assume that the values reported are the
336: central densities of the clusters. The total cluster mass and central
337: density and the calculated value of $b$ for each cluster are presented
338: in Table~\ref{table-props}.
339:
340: \subsection{Wind Sources}
341:
342: We implement wind sources as literal sources of SPH particles, using
343: the scheme described in \citet{Rock04}. The mass loss rates and wind
344: velocities inferred from observations are reported in
345: Tables~\ref{Arches.srcs} and \ref{Quint.srcs}. We position each
346: source at its observed $x$ and $y$ location and choose the $z$
347: coordinate randomly, subject to the constraint that the wind sources
348: are uniformly distributed over a range in $z$ equal to the observed
349: range in $x$ and $y$. The choice of $z$ positions has a much smaller
350: effect on the X-ray luminosity than the choice of mass-loss rate
351: (discussed below); \citet{Rock04} performed simulations of wind
352: sources in the central few parsecs of the Galaxy with two different
353: sets of $z$ positions, and the average $2-10$~keV X-ray luminosity
354: from the central 10$^{\prime\prime}$ of the simulations differed by
355: only 16\% ($7.50\times10^{31}$~erg~s$^{-1}$~arcsec$^{-2}$ from a
356: simulation with a dense arrangement of wind sources in the center of
357: the volume of solution versus
358: $6.45\times10^{31}$~erg~s$^{-1}$~arcsec$^{-2}$ from a simulation with
359: wind sources uniformly distributed in $z$). In addition, \citet{R01}
360: conducted three simulations of the Arches cluster in which $z$
361: positions of sources were chosen by sampling from a distribution
362: function $f(R)\propto R^{-2}$; they found a difference in $0.5-8$~keV
363: X-ray luminosity of only 3\% between the most and least luminous
364: simulations.
365:
366: Figure~\ref{fig-arches} shows the positions in the sky plane of the
367: wind sources in the Arches cluster, while Figure~\ref{fig-quint} shows
368: the wind sources in the Quintuplet cluster; the size of the circle
369: marking each source corresponds to the relative mass loss rate (on a
370: linear scale) for that star. The initial temperature of the winds is
371: not well known; for simplicity, we assume that all of the winds have a
372: temperature of $10^{4}$~K. Our results are fairly insensitive to this
373: value, however, since the temperature of the shocked gas is determined
374: primarily by the kinetic energy flux carried into the collision by the
375: winds. The sources are assumed to be stationary over the duration of
376: the simulation.
377:
378: We conduct two simulations of the wind-wind interactions in each
379: cluster; the simulations differ only in the choice of mass-loss rate
380: for each star, with the wind speed assumed constant. The ``standard''
381: simulations use mass-loss rates inferred from observations as
382: described in \S\S~\ref{sect.Arches} and \ref{sect.Quintuplet}, while
383: the ``high-\mdot'' simulations use mass-loss rates increased from the
384: standard value by a factor 2.
385:
386: \section{Observations}\label{sect.obs}
387:
388: An on-axis \textit{Chandra} observation of the Arches cluster was
389: taken on 2004 June 8 for 98.6~ksec. The ACIS-I detector was placed
390: at the focal plane. The data were analyzed with the latest CIAO
391: (version 3.1). While results of this observation will be presented in
392: an upcoming paper \citep{W04}, we concentrate here on the comparison
393: of the data with the simulations. Figure~\ref{chandra-arches} shows
394: an overlay of an X-ray intensity contour map on a \textit{HST} NICMOS
395: image of the Arches cluster \citep{F99b}. The actual spatial
396: resolution (with a FWHM $\lesssim 1^{\prime\prime}$) is better than
397: what appears on this smoothed image. The position coincidence of the
398: point-like X-ray sources and bright near-IR objects is apparent. To
399: compare with the simulated cluster wind properties, we remove a region
400: of twice the 90\% energy-encircled radius around each of the sources.
401:
402: \section{Results}\label{sect.results}
403:
404: All four simulations were run significantly past the point in time
405: when the stellar winds fill the volume of solution and gas shocked in
406: wind-wind collisions fills the core of each cluster. The Arches
407: cluster simulations were run past $t=10,000$~yr; the Quintuplet
408: simulations were run for over $14,000$~yr. The winds in each
409: simulation reach the edge of the volume of solution after $\sim
410: 3,000$~yr, but the most relevant timescale for determining when the
411: simulations reach a steady level of X-ray luminosity is the time
412: required to fill the core of each cluster with shocked gas. The
413: Arches cluster core is roughly five times smaller in radius than the
414: core of the Quintuplet cluster, so the X-ray luminosity from the
415: Arches simulations reaches a steady value relatively quickly compared
416: to the Quintuplet simulations.
417:
418: \subsection{Total Flux and Time Variation}
419:
420: In order to calculate the observed continuum spectrum, we assume that
421: the observer is positioned along the positive $z$-axis at infinity and
422: we sum the emission from all of the gas injected by the stellar wind
423: sources into the volume of each calculation. For the conditions
424: encountered in the two clusters, scattering is negligible and the
425: optical depth is always less than unity. For these temperatures and
426: densities, the dominant components of the continuum emissivity are
427: electron-ion ($\epsilon_{ei}$) and electron-electron ($\epsilon_{ee}$)
428: bremsstrahlung. We assume that the gas is in ionization equilibrium,
429: although spectra obtained from the Arches cluster with
430: \textit{Chandra} indicate the presence of line emission at $6.4$~keV;
431: however, no other significant lines are observed, so our use of a
432: bremsstrahlung model without line emission is a reasonable
433: approximation. Future, more refined versions of the calculations
434: reported here will need to include the effects of partial ionization
435: in the shocked gas, a condition suggested by the iron line emission.
436: We note, however, that the overall energetics and diffuse X-ray
437: luminosity calculated by assuming full ionization err only marginally
438: since the fraction of charges free to radiate via bremsstrahlung
439: should be very close to one.
440:
441: The X-ray luminosities calculated from our simulations support the
442: recent result of \citet{L04} that the majority of the X-ray emission
443: from these clusters (e.g., $\sim 60\%$ for Arches) is probably due to
444: point sources. The diffuse 0.2--10~keV X-ray luminosity calculated
445: from our ``standard'' Arches cluster simulation---$5.4\times
446: 10^{34}$~erg~s$^{-1}$---falls below the 0.2--10~keV luminosity of
447: $4.1\times 10^{35}$~erg~s$^{-1}$ from the A1 and A2 components
448: reported in \citet{Y02}; the simulation using elevated estimates for
449: the mass-loss rates produces $2.2\times 10^{35}$~erg~s$^{-1}$ between
450: 0.2 and 10~keV, or 53\% of the emission observed by \textit{Chandra}.
451: On the other hand, \citet{L04} now identify A1 and A2 as point-like
452: components; after subtracting the contributions of A1 and A2,
453: \citet{Y02} find that the A3 component has a 0.5--10~keV luminosity of
454: $\sim 1.6\times 10^{34}$~erg~s$^{-1}$. Our ``standard'' simulation
455: produces $2.7\times10^{34}$~erg~s$^{-1}$ between 0.5 and 10~keV;
456: lowering the mass-loss estimates of all wind sources by $\sim 30\%$ or
457: slightly decreasing the assumed wind velocities would produce even
458: better agreement.
459:
460: Our ``standard'' and ``high-\mdot'' simulations of the Quintuplet
461: cluster produce $1.5\times10^{33}$~erg~s$^{-1}$ and
462: $5.9\times10^{33}$~erg~s$^{-1}$, respectively, between 0.5 and 8~keV.
463: \citet{L04} estimate a 0.5--8~keV luminosity of $\sim
464: 1\times10^{34}$~erg~s$^{-1}$ for the diffuse emission from the
465: Quintuplet cluster, but they point out that other regions of diffuse
466: emission to the north and south of the cluster introduce a complicated
467: pattern of background emission and limit the precision of this
468: estimate.
469:
470: The long-term variation of the X-ray luminosity as a function of time
471: demonstrates some of the key differences between the two clusters.
472: Figures~\ref{lvst-arches} and \ref{lvst-quint} show the time variation
473: of the 0.5--8~keV X-ray luminosity from all four simulated clusters.
474: Each large graph shows the variation over the course of the entire
475: simulation ($> 10,000$~yr for the simulations of the Arches cluster,
476: and $> 14,000$~yr for the Quintuplet simulations), while each inset
477: graph shows variation over $1,000$ timesteps, or $\sim 150$~yr. The
478: fact that the Quintuplet cluster is nearly five times larger in radius
479: than the Arches cluster means that much more time is required for
480: shocked gas to fill the central region. The Arches cluster reaches a
481: steady X-ray luminosity after less than $2,000$~yr, while the
482: luminosity of the Quintuplet cluster does not clearly level off until
483: more than $10,000$~yr have passed.
484:
485: The two clusters also exhibit differences in the size of short-term
486: fluctuations in X-ray luminosity. Both simulations of the Arches
487: cluster show short-term variations in luminosity of $\sim 1\%$ over
488: timescales of $\sim 50$~yr, while the X-ray luminosity in the
489: ``standard'' Quintuplet simulation fluctuates by $\sim 4\%$, and
490: short-term variations in the ``high-\mdot'' simulation of the
491: Quintuplet cluster are as large as $\sim 7\%$.
492:
493: \subsection{Spatial Variation of the X-ray Flux}
494:
495: Figure~\ref{chandra-arches} shows that many of the bright X-ray peaks
496: in the Arches cluster correspond to actual stars, presumably binaries
497: whose binary wind interactions produce the strong localized X-ray
498: emission. These sources must be subtracted to study the diffuse X-ray
499: emission.
500:
501: The simulated 0.5--8~keV X-ray contours from the region near the core
502: of the Arches cluster, shown in Figure~\ref{contour-arches}, are
503: generally comparable to the contours generated from \textit{Chandra}
504: observations \citep[Figure~\ref{chandra-arches}; see also Figure
505: 2b,][]{Y02}. The strongest emission in the simulations and in
506: \textit{Chandra} images comes from the core of the cluster, and both
507: sets of contours form elliptical patterns aligned primarily
508: north-south. X-ray contours from the simulation of the Quintuplet
509: cluster are shown in Figure~\ref{contour-quint}; the Quintuplet
510: cluster is significantly less dense than the Arches cluster, so the
511: X-ray emission is correspondingly less strongly peaked toward the
512: center of the cluster.
513:
514: The plots in Figure~\ref{lvsr} show the total 0.5--8~keV luminosity
515: from concentric columns aligned along the line of sight and extending
516: outward in radius from the center of the Arches (left) and Quintuplet
517: (right) clusters. The most luminous gas in the Arches cluster is
518: confined to within $20^{\prime\prime}$ of the center of the cluster;
519: in contrast, we include stars beyond $50^{\prime\prime}$ in the
520: Quintuplet cluster, and its luminosity continues to increase even
521: beyond a radius of $50^{\prime\prime}$.
522:
523: Similarly, the plots in Figure~\ref{fvsr} show the 2--8~keV X-ray flux
524: per square arcmin as a function of distance from the center of each
525: cluster from all four simulations. The individual crosses and error
526: bars in the graph from the Arches cluster represent flux measurements
527: from \textit{Chandra} observations of the cluster, after point sources
528: have been removed and an estimate of the background X-ray flux has
529: been subtracted. Here we assume that all of the emission centered
530: around NICMOS stellar sources is due to binary wind interactions. The
531: estimated background---0.064~counts~s$^{-1}$~arcmin$^{-2}$---is the
532: average number of counts obtained in an annulus between radii of
533: $0.5^{\prime}$ and $0.9^{\prime}$. The simulations of the Arches
534: cluster apparently produce more X-rays near the center of the cluster
535: but decrease in intensity more rapidly toward larger radii. The
536: relatively flat surface brightness profile evident in the
537: \textit{Chandra} data outside a radius of $0.3^{\prime}$ may arise in
538: part from the presence of additional background X-ray emission near
539: the Galactic center. It may also indicate confinement of the
540: X-ray-emitting gas in the Arches cluster by ram pressure exerted by a
541: molecular cloud or other external medium surrounding the cluster; our
542: calculations include no such medium, so the simulated gas escapes and
543: cools more rapidly as it leaves the core of the cluster.
544:
545: One way to constrain the amount of confining material surrounding the
546: cluster is to measure expansion velocities of the gas out of the
547: cluster. In our simulations, which did not include confining gas, the
548: material accelerates until it reaches an asymptotic velocity limit
549: roughly equal to the mean wind velocity (Figure~\ref{vvsr}). In the
550: Arches cluster, we reach that limit. In the (larger) Quintuplet
551: cluster, that limit apparently occurs at a distance from the cluster
552: center that is larger than the size of the simulation. However, if
553: molecular clouds or additional stars with strong winds are producing a
554: confining ram pressure around this cluster, the outflowing gas will
555: decelerate. Measurements of this velocity will constrain the
556: parameters of the surrounding material; future simulations can use
557: these constraints to include the effects of this material.
558:
559: \section{Discussion}\label{sect.discussion}
560:
561: Although the diffuse X-ray flux in clusters may be used to probe
562: stellar mass-loss, two notable complications in the case of the Arches
563: and Quintuplet clusters are the X-ray background present in the
564: Galactic center, and the contributions made to the overall emission by
565: point (i.e., binary wind) sources. The contribution made by the X-ray
566: background is difficult to quantify; \citet{L04} report that
567: background contributions lead to significant uncertainty in the
568: measured X-ray flux from the Quintuplet cluster. The contribution
569: from point sources is easier to handle; with its relatively high
570: spatial resolution, \textit{Chandra} can produce reasonable images in
571: which the required point-source subtraction may be made. Point
572: sources in the Arches cluster all exhibit a 6.7~keV line; the absence
573: of such a line in the spectrum of the diffuse emission indicates that
574: the point-source contribution to the diffuse emission is not likely to
575: be significant.
576: %However, the fact that our simulations produce an elongated profile
577: %for the Arches cluster without the inclusion of any point X-ray
578: %sources suggests that at least some of the X-rays from the strong
579: %sources A1 and A2 are due to diffuse---rather than point---emission.
580:
581: If, based on the observations made by \citet{Y02}, we assume that A1
582: and A2 are not point-like and do not subtract the contributions of
583: point sources from the overall X-ray flux, it appears that the diffuse
584: X-ray flux is consistent with higher mass-loss rates than the
585: ($0.3\times 10^{-5}\;M_\odot$~yr$^{-1}$) $3\times
586: 10^{-5}\;M_\odot$~yr$^{-1}$ assumed for stars (less than) above
587: 60\,M$_\odot$. As we have seen, however, the observed
588: point-source-subtracted diffuse emission matches our calculated X-ray
589: flux to within a factor 2 when we adopt the currently accepted stellar
590: mass-loss rates. Indeed, lowering the mass-loss rate estimates of all
591: wind sources by about $30\%$ would produce significant agreement
592: between theory and observation. But we must make sure that we have a
593: complete accounting of all the wind sources and carry through with a
594: more careful point-source subtraction before we can completely confirm
595: such claims.
596:
597: The fact that the simulated X-ray flux density from the Arches cluster
598: drops off more rapidly than the observed profile (see
599: Figure~\ref{fvsr}) means that (1) we may have underestimated the
600: contribution of the X-ray background; or, (2) we have ignored the
601: possibly important dynamical influence of a confining molecular gas
602: outside the cluster. Interpreting data beyond these radii (roughly
603: $0.3^{\prime}$ for Arches and $1.0^{\prime}$ for Quintuplet) requires
604: more detailed information on the cluster environment there.
605:
606: X-rays do prove to be an ideal probe of bulk mass-loss rates in
607: clusters. The X-ray emission depends sensitively on the mass-loss
608: rates of the constituent stars and, based on our simulations of the
609: Arches cluster, we can already limit the mass-loss rates to within a
610: factor of 2 of the currently accepted values. This also implies that
611: the assumed abundances in the Galactic center environment are
612: essentially correct, since the mass-loss rates from stellar models
613: depend on the adopted metallicity. In addition, studying the X-ray
614: emission from the outer region of the cluster may eventually lead to a
615: better understanding of the medium within which the cluster is
616: embedded. This is clearly relevant to the question of how these
617: unusual clusters came to be, and the relative roles played by
618: ``standard star formation'' versus stellar capture from outside the
619: cluster. The inferred stellar constituents of these clusters
620: \citep[e.g.,]{C96,F99a} also seem to be consistent with the required
621: mass loss rates, so the inferred unusual mass function of the Arches,
622: Quintuplet, and Central clusters continues to pose a challenge to our
623: understanding of how these stellar aggregates were first assembled.
624:
625: {\bf Acknowledgments} This research was partially supported by NASA
626: grant NAG5-9205 and NSF grant AST-0402502 at the University of
627: Arizona, and has made use of NASA's Astrophysics Data System Abstract
628: Service. FM is grateful to the University of Melbourne for its
629: support (through a Sir Thomas Lyle Fellowship and a Miegunyah
630: Fellowship). This work was also funded under the auspices of the U.S.
631: Dept. of Energy, and supported by its contract W-7405-ENG-36 to Los
632: Alamos National Laboratory and by a DOE SciDAC grant number
633: DE-FC02-01ER41176. The simulations were conducted primarily on the
634: Space Simulator at Los Alamos National Laboratory. This research used
635: resources of the National Energy Research Scientific Computing Center,
636: which is supported by the Office of Science of the U.S. Department of
637: Energy under Contract No. DE-AC03-76SF00098.
638: \clearpage
639: \begin{thebibliography}{}
640: \bibitem[Cant\'{o} et al.(2000)]{C00}Cant\'{o}, J., Raga, A. C., \&
641: Rodr\'{i}guez, L. F. 2000, \apj, 536, 896.
642: \bibitem[Cotera et al.(1996)]{C96}Cotera, A. S., Erickson, E. F.,
643: Colgan, S. W. J., Simpson, J. P., Allen, D. A., \& Burton, M. G.
644: 1996, \apj, 461, 750.
645: \bibitem[Figer et al.(1998)]{F98}Figer, D. F., Najarro, F., Morris,
646: M., McLean, I. S., Geballe, T. R., Ghez, A. M., \& Langer, N. 1998,
647: \apj, 506, 384.
648: \bibitem[Figer et al.(1999a)]{F99a}Figer, D. F., McLean, I. S., \&
649: Morris, M. 1999a, \apj, 514, 202.
650: \bibitem[Figer et al.(1999b)]{F99b}Figer, D. F., Kim, S. S., Morris,
651: M., Serabyn, E., Rich, R. M., \& McLean, I. S. 1999b, \apj, 525, 750.
652: \bibitem[Figer et al.(2002)]{F02}Figer, D. F., Najarro, F., Gilmore,
653: D., Morris, M., Kim, S. S., Serabyn, E., McLean, I. S., Gilbert,
654: A. M., Graham, J. R., Larkin, J. E., Levenson, N. A. \& Teplitz,
655: H. I. 2002, \apj, 581, 258.
656: \bibitem[Fryer \& Warren(2002)]{FW02}Fryer, C. L. \& Warren, M. S.
657: 2002, \apjl, 574, L65.
658: \bibitem[Gerhard(2001)]{G01}Gerhard, O. 2001, \apj, 546, L39.
659: \bibitem[Lang et al.(1999)]{L99}Lang, C. C., Figer, D. F., Goss, W.
660: M., \& Morris, M. 1999, \aj, 118, 2327.
661: \bibitem[Lang et al.(2001)]{L01}Lang, C. C., Goss, W. M., \&
662: Rodr\'{i}quez, L. F. 2001, \apjl, 551, L143.
663: \bibitem[Lang(2002)]{L02}Lang, C. C. 2002, in A Massive Star Odyssey,
664: from Main Sequence to Supernova, IAU Symp. 212, Eds. K. A. van der
665: Hucht, A. Herrero, \& C. Esteban.
666: \bibitem[Law \& Yusef-Zadeh(2004)]{L04}Law, C. \& Yusef-Zadeh, F.
667: 2004, \apj, accepted.
668: \bibitem[McMillan \& Portegies Zwart(2003)]{M03}McMillan, S. L. W. \&
669: Portegies Zwart, S. F. 2003, \apj, 596, 314.
670: \bibitem[Melia \& Falcke(2001)]{MF01}Melia, F. \& Falcke, H.
671: 2001, \araa, 39, 309.
672: \bibitem[Morris(1993)]{M93}Morris, M. 1993, \apj, 408, 496.
673: \bibitem[Nagata et al.(1995)]{N95}Nagata, T., Woodward, C. E., Shure,
674: M., \& Kobayashi, N. 1995, \aj, 109, 1676.
675: \bibitem[Ozernoy et al.(1997)]{OGU97}Ozernoy, L. M.,
676: Genzel, R., and Usov, V. V. 1997, \mnras, 288, 237.
677: \bibitem[Panagia \& Felli(1975)]{P75}Panagia, N. \& Felli, M. 1975,
678: \aap, 39, 1.
679: \bibitem[Plummer(1911)]{P11}Plummer, H. C. 1911, \mnras, 71, 460.
680: \bibitem[Raga et al.(2000)]{R00}Raga, A. C., Navarro-Gonz\'{a}lez, R.,
681: \& Villagr\'{a}n-Muniz, M. 2000, Rev. Mexicana Astron. Astrofis.,
682: 36, 67.
683: \bibitem[Raga et al.(2001)]{R01}Raga, A. C., Vel\'{a}zquez, P. F.,
684: Cant\'{o}, J., Masciadri, E., \& Rodr\'{i}guez, L. F. 2001, \apjl,
685: 559, L33.
686: \bibitem[Rockefeller et al.(2004)]{Rock04}Rockefeller, G., Fryer, C.
687: L., Melia, F., \& Warren, M. S. 2004, \apj, 604, 662.
688: \bibitem[Serabyn et al.(1998)]{S98}Serabyn, E., Shupe, D.,
689: \& Figer, D. F. 1998, \nat, 394, 448.
690: \bibitem[Wang(2003)]{W03} Wang, Q. D. 2003, IAU Symposium No. 214:
691: High Energy Processes and Phenomena in Astrophysics, Eds: X.D. Li,
692: V. Trimble, \& Z.R. Wang. 214, 32
693: \bibitem[Wang(2004)]{W04} Wang, Q. D. 2004, in preparation.
694: \bibitem[Warren et al.(2004)]{War04}Warren, M. S., Rockefeller, G., \&
695: Fryer, C. L. 2004, in preparation.
696: \bibitem[Wellstein \& Langer(1999)]{WL99} Wellstein, S. \& Langer,
697: N. 1999, A\&A, 350, 148
698: \bibitem[Yusef-Zadeh et al.(2002)]{Y02}Yusef-Zadeh, F., Law, C.,
699: Wardle, M., Wang, Q. D., Fruscione, A., Lang, C. C., \& Cotera, A.
700: 2002, \apj, 570, 665.
701: \end{thebibliography}{}
702:
703: %% 7/2/2004: Arches table data is correct (checked against
704: %% ~/src/init/wind/arches-fullsources.dat (for positions) and
705: %% arches-newsources.dat (for mass loss rates)).
706:
707: \clearpage
708: \begin{deluxetable}{lccccc}
709: \tablewidth{0pt}
710: \tablecaption{Parameters for the Arches Cluster Wind Sources\label{Arches.srcs}}
711: \tablehead{
712: \colhead{Star\tablenotemark{a}}
713: & \colhead{x\tablenotemark{b}}
714: & \colhead{y\tablenotemark{b}}
715: & \colhead{z\tablenotemark{c}}
716: & \colhead{v}
717: & \colhead{\mdot} \\
718:
719: & \colhead{(arcsec)}
720: & \colhead{(arcsec)}
721: & \colhead{(arcsec)}
722: & \colhead{(km s$^{-1}$)}
723: & \colhead{(${10}^{-5}\;M_\odot$ yr$^{-1}$)}
724: }
725: \startdata
726:
727: 1, \phn AR3 & \phn\phs0.00 & \phn\phs0.00 & \phn\phs8.38 & 1,000 & \phn3.2 \\
728: 2 & \phn$-$6.75 & \phn$-$3.53 & \phs10.70 & 1,000 & \phn3.0 \\
729: 3, \phn AR7 & \phn\phs8.20 & \phn$-$4.13 & \phn\phs2.66 & 1,000 & \phn4.2 \\
730: 4, \phn AR5 & \phn\phs4.83 & \phn\phs4.66 & \phn\phs2.71 & 1,000 & \phn3.0 \\
731: 5, \phn AR8 & \phn\phs3.29 & \phn$-$9.64 & \phn$-$5.31 & 1,000 & \phn3.6 \\
732: 6, \phn AR1 & \phn\phs2.87 & \phn$-$0.03 & \phn$-$1.85 & 1,000 & 17.0 \\
733: 7, \phn AR4 & \phn\phs3.53 & \phn\phs2.73 & \phn$-$5.72 & 1,000 & \phn3.9 \\
734: 8, \phn AR2 & \phn\phs2.46 & \phn\phs1.01 & \phn\phs2.78 & 1,000 & \phn3.9 \\
735: 9 & \phn\phs0.80 & \phs10.50 & \phn\phs0.23 & 1,000 & \phn3.0 \\
736: 10 & \phn$-$1.83 & \phn$-$4.25 & \phn$-$0.90 & 1,000 & \phn3.0 \\
737: 11 & \phn$-$1.03 & \phs14.41 & \phn\phs6.83 & 1,000 & \phn3.0 \\
738: 12 & \phn\phs1.01 & \phn\phs4.98 & $-$11.46 & 1,000 & \phn3.0 \\
739: 13 & \phn$-$2.08 & \phn$-$1.39 & \phn\phs6.12 & 1,000 & \phn3.0 \\
740: 14 & \phn\phs6.24 & \phn$-$0.32 & \phn\phs5.15 & 1,000 & \phn3.0 \\
741: 15 & \phn\phs7.24 & \phn\phs5.67 & $-$14.94 & 1,000 & \phn3.0 \\
742: 16 & \phn\phs4.22 & \phn\phs1.59 & $-$14.62 & 1,000 & \phn3.0 \\
743: 17 & \phn$-$0.89 & \phn$-$4.90 & \phs14.59 & 1,000 & \phn3.0 \\
744: 18, AR9 & \phn\phs3.58 & \phn\phs4.34 & \phs13.79 & 1,000 & \phn3.2 \\
745: 19, AR6 & \phn$-$5.81 & \phn$-$3.72 & \phn$-$4.97 & 1,000 & \phn4.5 \\
746: 20 & \phn\phs2.90 & \phn\phs2.58 & \phn$-$3.20 & 1,000 & \phn3.0 \\
747: 21 & \phn\phs7.36 & \phn\phs2.65 & \phs10.49 & 1,000 & \phn3.0 \\
748: 22 & \phn\phs0.24 & \phn\phs5.55 & \phn$-$8.01 & 1,000 & \phn3.0 \\
749: 23 & \phs12.50 & \phn$-$1.08 & \phs14.43 & 1,000 & \phn3.0 \\
750: 24 & \phn$-$1.42 & \phn\phs1.55 & \phn\phs9.62 & 1,000 & \phn3.0 \\
751: 25 & \phn$-$3.26 & \phn$-$4.30 & \phn$-$8.37 & 1,000 & \phn3.0 \\
752: 26 & \phn\phs4.60 & \phn$-$1.27 & \phs13.43 & 1,000 & \phn3.0 \\
753: 27 & \phn\phs5.31 & \phn\phs2.74 & \phn\phs4.80 & 1,000 & \phn3.0 \\
754: 28 & \phn\phs5.77 & \phn\phs0.55 & \phs10.79 & 1,000 & \phn3.0 \\
755: 29 & \phn\phs7.08 & \phn\phs4.62 & \phs11.95 & 1,000 & \phn3.0 \\
756: 36 & \phn$-$6.19 & \phs14.87 & \phn$-$8.33 & 1,000 & \phn3.0 \\
757: 37 & \phn\phs3.54 & \phn\phs2.99 & \phn\phs5.89 & 1,000 & \phn3.0 \\
758: 49 & \phn$-$1.74 & \phs14.97 & \phn\phs9.01 & 1,000 & \phn0.3 \\
759: 61 & \phn$-$1.53 & \phs23.67 & \phn\phs2.02 & 1,000 & \phn0.3 \\
760: 75 & \phn\phs7.42 & \phs11.51 & \phn\phs4.86 & 1,000 & \phn0.3 \\
761: 108 & \phn\phs7.22 & \phs11.83 & \phn\phs7.08 & 1,000 & \phn0.3 \\
762: 111 & \phn\phs0.65 & \phs18.90 & $-$12.39 & 1,000 & \phn0.3 \\
763: 116 & \phn\phs3.64 & \phs16.48 & \phs12.31 & 1,000 & \phn0.3 \\
764: 126 & \phn\phs8.80 & \phs19.13 & \phn$-$0.92 & 1,000 & \phn0.3 \\
765: 129 & \phn$-$9.61 & \phs10.04 & \phn\phs2.44 & 1,000 & \phn0.3 \\
766: 132 & \phn\phs7.04 & \phs20.08 & \phn\phs6.02 & 1,000 & \phn0.3 \\
767: 149 & \phn\phs5.54 & \phs21.20 & \phn$-$5.47 & 1,000 & \phn0.3 \\
768: 156 & \phn\phs5.02 & \phs20.61 & \phn\phs7.93 & 1,000 & \phn0.3 \\
769:
770: \enddata
771:
772: \tablenotetext{a}{Numerical designations taken from \citet{F02};
773: ``AR'' designations taken from \citet{L02}.}
774:
775: \tablenotetext{b}{Offset from $\alpha(2000)$: $17^h\,45^m\,50.26^s$,
776: $\delta(2000)$: $-28^\circ\,49^{\prime}\,22^{\prime\prime}.76$
777: \citep{F02}. Here, positive $x$ is ascending R.A. (to the East) and
778: positive $y$ is ascending declination (to the North).}
779:
780: \tablenotetext{c}{Simulated with Monte Carlo.}
781:
782: \end{deluxetable}
783: \clearpage
784:
785:
786: %% 7/2/2004: Quintuplet table data is correct (checked against
787: %% ~/src/init/wind/quint-withz.dat).
788:
789: \begin{deluxetable}{lccccc}
790: \tablewidth{0pt}
791: \tablecaption{Parameters for the Quintuplet Cluster Wind Sources\label{Quint.srcs}}
792: \tablehead{
793: \colhead{Star\tablenotemark{a}}
794: & \colhead{x\tablenotemark{b}}
795: & \colhead{y\tablenotemark{b}}
796: & \colhead{z\tablenotemark{c}}
797: & \colhead{v}
798: & \colhead{\mdot} \\
799:
800: & \colhead{(arcsec)}
801: & \colhead{(arcsec)}
802: & \colhead{(arcsec)}
803: & \colhead{(km s$^{-1}$)}
804: & \colhead{(${10}^{-5}\;M_\odot$ yr$^{-1}$)}
805: }
806: \startdata
807:
808: QR1 & \phn$-$3.30 & \phs20.10 & $-$36.41 & \phn\phm{,}500 & \phn8.0\phn \\
809: QR2 & \phn$-$0.30 & \phs21.80 & \phn\phs9.39 & \phn\phm{,}500 & 15.1\phn \\
810: QR3 & \phn\phs8.30 & \phs22.90 & $-$42.78 & \phn\phm{,}500 & \phn6.7\phn \\
811: QR6 & \phn$-$1.80 & \phn\phs9.00 & \phs31.80 & 1,000 & \phn6.1\phn \\
812: QR7 & \phn\phs3.00 & \phn\phs1.00 & $-$26.99 & 1,000 & 13.0\phn \\
813: 76 & \phn$-$9.00 & $-$36.60 & \phn$-$9.50 & 1,000 & \phn0.28 \\
814: 134, Pistol & \phn$-$4.50 & $-$21.90 & $-$28.26 & \phn\phm{,}100 & \phn3.8\phn \\
815: 151 & \phn\phs1.50 & $-$18.80 & $-$20.94 & 1,000 & \phn1.1\phn \\
816: 157 & \phs15.00 & $-$17.00 & \phn\phs8.82 & \phn\phm{,}500 & \phn0.16 \\
817: 235 & \phn$-$4.50 & \phn\phs1.50 & \phn$-$2.91 & 1,000 & \phn3.7\phn \\
818: 240 & $-$15.00 & \phn\phs3.80 & \phn$-$9.06 & 1,000 & \phn3.3\phn \\
819: 241, QR5 & \phn$-$3.00 & \phn\phs4.80 & \phs39.75 & 1,000 & \phn6.6\phn \\
820: 250 & \phn$-$7.50 & \phn\phs7.10 & \phs48.44 & \phn\phm{,}500 & \phn1.0\phn \\
821: 256 & $-$24.00 & \phn\phs9.80 & \phn$-$9.66 & 1,000 & \phn0.70 \\
822: 257 & \phn$-$4.50 & \phn\phs9.50 & $-$19.34 & \phn\phm{,}500 & \phn1.0\phn \\
823: 269 & \phn$-$9.00 & \phs11.90 & \phs24.33 & \phn\phm{,}500 & \phn0.1\phn \\
824: 270, QR4 & \phn$-$3.00 & \phs13.10 & $-$29.77 & \phn\phm{,}500 & \phn1.4\phn \\
825: 274 & $-$39.00 & \phs12.60 & \phs42.80 & 1,000 & \phn0.67 \\
826: 276 & \phs22.50 & \phs12.70 & \phn$-$7.87 & \phn\phm{,}500 & \phn0.1\phn \\
827: 278 & \phn$-$3.00 & \phn\phs7.50 & $-$36.93 & \phn\phm{,}500 & \phn1.0\phn \\
828: 301 & $-$16.50 & \phs20.00 & \phn$-$4.93 & \phn\phm{,}500 & \phn0.1\phn \\
829: 307 & \phn$-$9.00 & \phs21.50 & $-$41.15 & \phn\phm{,}500 & \phn1.0\phn \\
830: 309 & $-$39.00 & \phs22.90 & $-$38.32 & 1,000 & \phn0.26 \\
831: 311 & \phs18.00 & \phs22.50 & \phs34.69 & \phn\phm{,}500 & \phn0.1\phn \\
832: 320 & \phs12.00 & \phs25.30 & $-$18.66 & 1,000 & \phn0.61 \\
833: 344 & $-$27.00 & \phs32.40 & \phn\phs4.04 & \phn\phm{,}500 & \phn0.1\phn \\
834: 353 & \phs55.50 & \phs36.70 & \phn$-$6.15 & 1,000 & \phn0.22 \\
835: 358 & $-$25.50 & \phs36.80 & $-$15.30 & \phn\phm{,}500 & \phn0.32 \\
836: 362 & $-$46.50 & \phs38.40 & $-$13.75 & \phn\phm{,}100 & \phn3.0\phn \\
837: 381 & \phs21.00 & \phs42.80 & $-$22.20 & \phn\phm{,}500 & \phn0.40 \\
838: 406 & \phs15.00 & \phs51.50 & $-$26.93 & \phn\phm{,}500 & \phn0.1\phn \\
839:
840: \enddata
841:
842: \tablenotetext{a}{Numerical designations taken from \citet{F99a};
843: ``QR'' designations taken from \citet{L02}.}
844:
845: \tablenotetext{b}{Offset from $\alpha(1950)$: $17^h\,43^m\,4.5^s$,
846: $\delta(1950)$: $-28^\circ\,48^{\prime}\,35^{\prime\prime}$
847: \citep[based on][]{L99}. Here, positive $x$ is ascending R.A. (to
848: the East) and positive $y$ is ascending declination (to the North).}
849:
850: \tablenotetext{c}{Simulated with Monte Carlo.}
851:
852: \end{deluxetable}
853: \clearpage
854:
855: %% 7/8/2004: checked with F02 and personal notes
856:
857: \begin{deluxetable}{lcccc}
858: \tablewidth{0pt}
859: \tablecaption{Properties of the Arches and Quintuplet Clusters\label{table-props}}
860: \tablehead{
861: \colhead{Cluster}
862: & \colhead{M\tablenotemark{a}}
863: & \colhead{$\rho_{0}$\tablenotemark{a}}
864: & \colhead{b} \\
865:
866: & \colhead{($M_{\odot}$)}
867: & \colhead{($M_{\odot}$ pc$^{-3}$)}
868: & \colhead{(pc)}
869: }
870: \startdata
871:
872: Arches & $2.0\times 10^{4}$ & $6.3\times 10^{5}$ & 0.20 \\
873: Quintuplet & $6.3\times 10^{3}$ & $1.6\times 10^{3}$ & 0.98 \\
874:
875: \enddata
876:
877: \tablenotetext{a}{See \citet{F99a}.}
878: \end{deluxetable}
879: \clearpage
880: \begin{figure}
881: \epsscale{1.00}
882: \plotone{f1.eps}
883: \caption{Location of the 42 wind-producing stars used in the
884: simulations of the Arches cluster, relative to $\alpha(2000)$:
885: $17^h\,45^m\,50.26^s$, $\delta(2000)$:
886: $-28^\circ\,49^{\prime}\,22^{\prime\prime}.76$ \citep{F02}. The radius
887: of each circle corresponds (on a linear scale) to that star's mass
888: loss rate. Setting the scale is AR1, with $\dot M=1.7\times
889: 10^{-4}\;M_\odot$ yr$^{-1}$.}
890: \label{fig-arches}
891: \end{figure}
892: \clearpage
893: \begin{figure}
894: \epsscale{1.00}
895: \plotone{f2.eps}
896: \caption{Location of the 31 wind-producing stars used in the
897: simulations of the Quintuplet cluster, relative to $\alpha(1950)$:
898: $17^h\,43^m\,4.5^s$, $\delta(1950)$:
899: $-28^\circ\,48^{\prime}\,35^{\prime\prime}$ \citep{L99}. The radius
900: of each circle corresponds (on a linear scale) to that star's mass
901: loss rate. Setting the scale is QR2, with $\dot M=1.5\times
902: 10^{-4}\;M_\odot$ yr$^{-1}$.}
903: \label{fig-quint}
904: \end{figure}
905: \clearpage
906: \begin{figure}
907: \epsscale{1.00}
908: \includegraphics[angle=90]{f3.eps}
909: \caption{\textit{Chandra} ACIS-I 1--9 keV intensity contours overlaid on a
910: \textit{HST} NICMOS image of the Arches cluster. The image is
911: exposure-corrected and is adaptively smoothed with the CIAO csmooth
912: routine ($S/N \sim 3\sigma$). The contour levels are at $(31, 32,
913: 34, 38, 46, 62, 94, 158, 286, 542, 1054, 2078, 4126) \times 10^{-3}
914: {\rm~counts~s^{-1}~arcmin^{-2}}$.}
915: \label{chandra-arches}
916: \end{figure}
917: \clearpage
918: \begin{figure}
919: \epsscale{1.00}
920: \plottwo{f4a.eps}{f4b.eps}
921: \caption{The 0.5--8 keV X-ray luminosity versus time from the ``standard''
922: (left) and ``high-\mdot'' simulations of the Arches cluster. The
923: large plot shows the variation in luminosity over the entire
924: calculation, while the inset plot shows variation on a timescale of
925: $\sim$~10~years. The winds fill the core of the cluster after
926: $\sim$~2,000 years.}
927: \label{lvst-arches}
928: \end{figure}
929: \clearpage
930: \begin{figure}
931: \epsscale{1.00}
932: \plottwo{f5a.eps}{f5b.eps}
933: \caption{The 0.5--8 keV X-ray luminosity versus time from the ``standard''
934: (left) and ``high-\mdot'' simulations of the Quintuplet cluster.
935: The large plot shows the variation in luminosity over the entire
936: calculation, while the inset plot shows variation on a timescale of
937: $\sim$~10~years. The winds fill the core of the cluster after
938: $\sim$~10,000 years.}
939: \label{lvst-quint}
940: \end{figure}
941: \clearpage
942: \begin{figure}
943: \epsscale{1.00}
944: \plottwo{f6a.eps}{f6b.eps}
945: \caption{Contours of column-integrated 0.5--8 keV X-ray luminosity per
946: arcsec$^2$ from the ``standard'' (left) and ``high-\mdot''
947: simulations of the Arches cluster. In order from blue to cyan to
948: green to red, the luminosities indicated by the contours are
949: $10^{30}$, $2.5\times10^{30}$, $8\times10^{30}$, and
950: $2\times10^{31}$~ergs~s$^{-1}$~arcsec$^{-2}$. The crosses indicate
951: the positions of wind sources included in the simulations.}
952: \label{contour-arches}
953: \end{figure}
954: \clearpage
955: \begin{figure}
956: \epsscale{1.00}
957: \plottwo{f7a.eps}{f7b.eps}
958: \caption{Contours of column-integrated 0.5--8 keV X-ray luminosity per
959: arcsec$^2$ from the ``standard'' (left) and ``high-\mdot''
960: simulations of the Quintuplet cluster. In order from blue to cyan
961: to green to red, the luminosities indicated by the contours are
962: $7\times10^{28}$, $1.8\times10^{29}$, $7\times10^{29}$, and
963: $2\times10^{30}$~ergs~s$^{-1}$~arcsec$^{-2}$. The crosses indicate
964: the positions of wind sources included in the simulations.}
965: \label{contour-quint}
966: \end{figure}
967: \clearpage
968: \begin{figure}
969: \epsscale{1.00}
970: \plottwo{f8a.eps}{f8b.eps}
971: \caption{Total 0.5--8~keV X-ray luminosity from concentric cylinders
972: aligned with the center of the Arches (left) and Quintuplet
973: clusters. The lower line in each graph represents the ``standard''
974: simulation of that cluster; the upper line represents the
975: ``high-\mdot'' simulation.}
976: \label{lvsr}
977: \end{figure}
978: \clearpage
979: \begin{figure}
980: \epsscale{1.00}
981: \plottwo{f9a.eps}{f9b.eps}
982: \caption{The 2--8~keV X-ray flux per arcmin$^{2}$ from concentric
983: cylinders aligned with the center of the Arches (left) and
984: Quintuplet clusters. The lower line in each graph represents the
985: ``standard'' simulation of that cluster; the upper line represents
986: the ``high-\mdot'' simulation. Crosses and error bars in the left
987: graph represent flux measurements from \textit{Chandra} observations
988: of the Arches cluster.}
989: \label{fvsr}
990: \end{figure}
991: \clearpage
992: \begin{figure}
993: \epsscale{1.00}
994: \plottwo{f10a.eps}{f10b.eps}
995: \caption{Average radial velocity of gas in the ``standard'' simulations of
996: the Arches (left) and Quintuplet clusters, versus distance from the
997: center of each cluster.}
998: \label{vvsr}
999: \end{figure}
1000:
1001: \end{document}
1002:
1003: