1: \pdfoutput=1
2: \documentclass[pdftex]{pasj00}
3:
4: \newcommand{\av}{\ensuremath{A_{V}}}
5: \newcommand{\ji}{\textit{J}}
6: \newcommand{\hi}{\textit{H}}
7: \newcommand{\ki}{\textit{K}}
8: \newcommand{\ksi}{\textit{K\ensuremath{_S}}}
9: \newcommand{\hii}{H\,$_{\rm{II}}$}
10: \newcommand{\kt}{\ensuremath{k_{\rm{B}}T}}
11: \newcommand{\lx}{\ensuremath{L_{\rm{X}}}}
12: \newcommand{\lbol}{\ensuremath{L_{\rm{bol}}}}
13: \newcommand{\fx}{\ensuremath{F_{\rm{X}}}}
14: \newcommand{\nh}{\ensuremath{N_{\rm H}}}
15:
16: \SetRunningHead{D.~Takei et al.}{Detection of an X-Ray Transient Source}
17: \Received{2007/04/04}
18: \Accepted{2007/05/16}
19: \Published{}
20:
21: \begin{document}
22:
23: \title{Detection of a Rare Supersoft Outburst Event \\ during a Suzaku Observation of 1E\,0102.2--7219}
24:
25: \author{
26: Dai \textsc{Takei}\altaffilmark{1},
27: Masahiro \textsc{Tsujimoto}\altaffilmark{1,2,3},
28: Shunji \textsc{Kitamoto}\altaffilmark{1},
29: Mikio \textsc{Morii}\altaffilmark{1},\\
30: Ken \textsc{Ebisawa}\altaffilmark{4},
31: Yoshitomo \textsc{Maeda}\altaffilmark{4},
32: and Eric D. \textsc{Miller}\altaffilmark{5}
33: }
34:
35: \altaffiltext{1}{Department of Physics, Rikkyo University\\
36: 3-34-1 Nishi-Ikebukuro, Toshima, Tokyo 171-8501}
37: \email{takei@stu.rikkyo.ne.jp}
38: \altaffiltext{2}{Department of Astronomy and Astrophysics, Pennsylvania State University\\
39: 525 Davey Laboratory, University Park, PA 16802, USA}
40: \altaffiltext{3}{Chandra Fellow}
41: \altaffiltext{4}{Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency\\
42: 3-1-1 Yoshinodai, Sagamihara, Kanagawa 229-8510}
43: \altaffiltext{5}{Massachusetts Institute of Technology, Kavli Institute for Astrophysics and Space Research\\
44: 77 Massachusetts Avenue 37-551, Cambridge, MA 02139, USA}
45:
46: \KeyWords{
47: stars: individual (Suzaku\,J0105--72) ---
48: stars: novae, cataclysmic variables ---
49: stars: white dwarfs
50: }
51:
52: \maketitle
53:
54: \begin{abstract}
55: We report the detection of a transient X-ray source toward the Small Magellanic Cloud
56: (SMC) using the X-ray Imaging Spectrometer (XIS) onboard the Suzaku telescope. The
57: source was detected at the edge of the XIS image during a routine observation of the
58: calibration source 1E\,0102.2--7219, a supernova remnant in the SMC. We constrained the
59: source position using ray-tracing simulations. No such transient source was found at
60: the position in the other Suzaku observations nor in all the available archived images
61: of other X-ray missions for the last $\sim$28 years. The XIS spectrum can be explained
62: by a single blackbody with a temperature of $\sim$72~eV, and an interstellar extinction
63: of $\sim$4.9$\times$10$^{20}$~H~atoms~cm$^{-2}$ consistent with the value to the
64: SMC. An additional absorption edge at $\sim$0.74~keV was also confirmed, which is
65: presumably due to the absorption by helium-like oxygen ions. Assuming that the source
66: is at the distance of the SMC, the X-ray luminosity in the 0.2--2.0~keV band is
67: $\sim$10$^{37}$~erg~s$^{-1}$ and the radius of the source is $\sim$10$^{8}$~cm. The XIS
68: light curve shows about a two-fold decline in X-ray flux during the 24~ks
69: observation. Together with the archived data, the X-ray flux in the burst is at least
70: three orders of magnitude brighter than the undetected quiescent level. All these
71: properties are often seen among supersoft sources (SSSs). We conclude that the
72: transient source is another example of SSS in the SMC.
73: \end{abstract}
74:
75: \section{Introduction}
76: 1E\,0102.2--7219 is the second brightest X-ray source in the Small Magellanic Cloud
77: (SMC) at a distance of $\sim$60~kpc. It was discovered by the Einstein satellite
78: \citep{seward81} and was found to be a shell-type supernova remnant with a radius of
79: $\sim$14\arcsec~\citep{gaetz00}. The source is particularly suited for the routine
80: calibration of the quantum efficiency, the energy gain and resolution of X-ray detectors
81: because of its soft and line-dominated spectrum, non-variable flux, and good visibility
82: from satellites throughout the year. Therefore, a large number of pointed observations
83: of this object was conducted by X-ray satellites including Einstein, ROSAT, ASCA,
84: Beppo-SAX, Chandra, XMM-Newton, and Suzaku (\cite{hayashi94}; \cite{gaetz00};
85: \cite{hughes00}; \cite{sasaki01}; \cite{rasmussen01}; \cite{flanagan04};
86: \cite{sasaki06}). This offers a unique opportunity to search for transient sources in
87: the surrounding area and monitor their long-term behavior without additional investments
88: of telescope times.
89:
90: Among sixteen Suzaku observations conducted by 2007 March, we detected a transient
91: source at the edge of the image obtained on 2005 August 31. As we discuss in this paper,
92: no such transient source was detected in any of the remaining fifteen Suzaku and
93: archived Einstein, ROSAT, ASCA, Beppo-SAX, Chandra, and XMM-Newton observations despite
94: their frequent visits, which indicates that the burst event of this source is quite
95: rare. Here, we report the result of the Suzaku observation of the transient source and
96: discuss its nature.
97:
98: \section{Observations}
99: Sixteen observations of 1E\,0102.2--7219 were performed using the Suzaku satellite by
100: 2007 March (table~\ref{tb:t1}). Suzaku \citep{mitsuda07} has two instruments in
101: operation; the X-ray Imaging Spectrometer (XIS; \cite{koyama07}) and the Hard X-ray
102: Detector (HXD; \cite{takahashi07}; \cite{kokubun07}). We concentrate on the XIS data in
103: this paper.
104:
105: The XIS is equipped with four X-ray charge coupled devices (CCDs) at the foci of four
106: X-Ray Telescopes (XRT; \cite{serlemitsos07}). Three of them (XIS0, 2, and 3) are
107: front-illuminated (FI) CCDs sensitive in the 0.4--12~keV energy range and the remaining
108: one (XIS1) is a back-illuminated (BI) CCD sensitive in 0.2--12~keV. The total effective
109: area is $\sim$1360~cm$^{2}$ at 1.5~keV. XIS covers a $\sim$18\arcmin$\times$18\arcmin\
110: field of view (FoV), with an energy-independent half power diameter of
111: 1\farcm8--2\farcm3. Each CCD has a format of 1024$\times$1024 pixels with a pixel scale
112: of $\sim$1\arcsec~pixel$^{-1}$. Two radioactive \atom{Fe}{}{55} sources illuminate two
113: corners of each CCD.
114:
115: The energy resolution and quantum efficiency of the XIS are gradually degrading in
116: orbit. Their trend in the soft band is monitored using 1E\,0102.2--7219 in the same data
117: sets presented here. As of 2005 August 31, when the transient source was detected, the
118: absolute energy scale is accurate to $\lesssim$5~eV, and the energy resolution is
119: $\sim$63~eV (FI) and $\sim$71~eV (BI) at 1.0~keV. Due to an unknown contaminant
120: accumulating on the optical blocking filters of the XIS, the effective area in the soft
121: band has been diminishing since shortly after launch \citep{koyama07}. This effect is
122: accounted for in the ancillary response files and ray-tracing simulators of the
123: telescopes \citep{ishisaki07}.
124:
125: All the XIS data were taken with the normal clocking mode with a frame time of 8~s. The
126: first two observations were aimed at (R.\,A., Dec.)\,$=$\,(\timeform{01h04m02s},
127: \timeform{-72D02'00''}), while the others were at (R.\,A.,
128: Dec.)\,$=$\,(\timeform{01h04m02s}, \timeform{-72D01'53''}) in the equinox J2000.0. We
129: used the data products of the processing version 1.2, in which events were removed
130: during the South Atlantic anomaly passages, the night earth elevation angle below
131: 5~degrees, or the day earth elevation angle below 20~degrees \citep{fujimoto07}.
132:
133: \begin{table}
134: \begin{center}
135: \caption{Suzaku observation log}\label{tb:t1}
136: \begin{tabular}{cllcc}
137: \hline\hline
138: \phantom{000} & \multicolumn{1}{c}{Sequence} & \multicolumn{1}{c}{Start} & $\textit{t}\rm{_{exp}}$\footnotemark[$*$] & \phantom{000} \\
139: & \multicolumn{1}{c}{number} & \multicolumn{1}{c}{date} & (ks) & \\
140: \hline
141: & 100001020 & 2005-08-13 & \phantom{0}4 & \\
142: & 100014010\footnotemark[$\dagger$] & 2005-08-31 & 24 & \\
143: & 100044010 & 2005-12-16 & 71 & \\
144: & 100044020 & 2006-01-17 & 42 & \\
145: & 100044030 & 2006-02-02 & 21 & \\
146: & 101005010 & 2006-04-16 & 22 & \\
147: & 101005020 & 2006-05-21 & 19 & \\
148: & 101005030 & 2006-06-26 & 22 & \\
149: & 101005040 & 2006-07-17 & 22 & \\
150: & 101005050 & 2006-08-25 & 49 & \\
151: & 101005060 & 2006-09-19 & 11 & \\
152: & 101005070 & 2006-10-21 & 37 & \\
153: & 101005090 & 2006-12-13 & 28 & \\
154: & 101005100 & 2007-01-15 & 24 & \\
155: & 101005110 & 2007-02-10 & 36 & \\
156: & 101005120 & 2007-03-18 & 18 & \\
157: \hline
158: \multicolumn{5}{@{}l@{}}{\hbox to 0pt{\parbox{80mm}{\footnotesize
159: \footnotemark[$*$] Averaged exposure time of the operating CCDs.
160: \par\noindent
161: \footnotemark[$\dagger$] The transient source was detected in this observation.
162: }\hss}}
163: \end{tabular}
164: \end{center}
165: \end{table}
166:
167: \section{Analysis}
168: \subsection{Image Analysis}
169: Figure~\ref{fg:f1} shows the XIS image obtained on 2005 August 31. Events taken with the
170: four detectors were merged. XIS detected four sources. Three of them were identified as
171: 1E\,0102.2--7219, RX\,J0103.6--7201, and 2E\,0101.5--7225 using the SIMBAD
172: database. These sources were detected in all images taken at the other epochs. The
173: remaining one is a transient source detected only in the soft-band image on 2005 August
174: 31. The position of the transient source is very close to the edge of the detector and
175: its image is truncated. In order to constrain the position as precisely as possible, we
176: took the following procedure. Throughout this paper, we used ray-tracing software
177: designed for the Suzaku telescopes (\texttt{xissim}; \cite{ishisaki07}) to simulate XIS
178: images. Simulation were conducted only for XIS1; the source is quite soft and most of
179: the events were recorded in XIS1.
180:
181: \begin{figure}
182: \begin{center}
183: \FigureFile(80mm,80mm){1a.jpg}
184: \FigureFile(80mm,80mm){1b.jpg}
185: \end{center}
186: \caption{
187: XIS images in the (a) 0.2--2.0~keV and (b) 2.0--12~keV bands. Events taken with the
188: four XIS were merged in the data on 2005 August 31. The bright spots in the four
189: corners in (b) are the calibration sources. The solid and the dashed circles indicate
190: the source and background accumulation regions, respectively. The names of the three
191: persistent sources are given. The XIS frame was registered to match the observed
192: position of 1E\,0102.2--7219 with the Chandra position.}\label{fg:f1}
193: \end{figure}
194:
195: First, we registered the astrometry of the XIS image by matching the position of
196: 1E\,0102.2--7219 with that of Chandra. The astrometry of the Chandra image is accurate
197: to $\sim$0\farcs6\footnote{See http://asc.harvard.edu/proposer/POG/ for details.}. We
198: simulated an XIS image based on the Chandra image (obsID\,$=$\,5139), and constructed
199: R.\,A. and Dec. projections of the surface brightness in the 0.2--2.0~keV band. The
200: observed profile was fitted to the simulated profile by shifting 13\arcsec\ and
201: 19\arcsec\ in the R.\,A. and Dec. directions, respectively. The uncertainty of the
202: fitting is estimated to be $\sim$1\arcsec\ in both directions, which is inherited to the
203: uncertainty of the position determination of the transient source.
204:
205: Second, we simulated XIS images of a point source at various positions toward the inside
206: and outside of the FoV from the observed peak of the source. Figure~\ref{fg:f2} shows
207: the layout of the simulated positions. We define the axes parallel and perpendicular to
208: the detector edge as $x$ and $y$, respectively. The origin of the $x$-axis is defined so
209: that the observed peak of the transient source is 0. The origin of the $y$-axis is
210: placed at the field edge, and negative and positive values of $y$ indicate that the
211: position is inside and outside of the field, respectively. The unit of the coordinate is
212: pixels. The apparent peak of the source is at ($x$, $y$)\,$=$\,(0, $-$17).
213:
214: We simulated XIS images at positions along the line between ($x$, $y$)\,$=$\,(0, $-$50)
215: and (0, 70) with a step of $\Delta y$\,$=$\,1 (the gray line in figure~\ref{fg:f2}). For
216: the simulated two-dimensional image at each position, we constructed a one-dimensional
217: surface brightness profile inside the FoV of the detector to match with the observed
218: profile. The profiles were projected on the $x$-axis and the counts in the range of
219: $y$~$>$~--183 were accumulated, so that it covers the entire extent of the observed
220: image. In the fitting, the simulated profile was allowed to move along the $x$-axis, so
221: that its peak matches with that of the observed profile.
222:
223: With the $x$-projected profile, we can constrain both the $x$ and $y$ positions of the
224: transient source. On one hand, the peak position of the profile is sensitive to the
225: assumed $x$ position of the transient source. On the other hand, the width of the
226: profile is sensitive to the assumed $y$ position; this is because the width of the
227: simulated profile becomes narrower as the assumed $y$ position moves away from the
228: edge. We do not use the $y$-projected profiles as their peaks and widths are insensitive
229: to the assumed position of the source.
230:
231: After this procedure for every point along the gray line in figure~\ref{fg:f2}, we
232: obtained the 90\% and 99\% confidence ranges of the source position, which are shown as
233: contours in figure~\ref{fg:f2}. We have two islands of local minimum; one (A in
234: figure~\ref{fg:f2}) is outside of the FoV at ($x$, $y$)\,$=$\,($-$8.5, 20) or (R.\,A.,
235: Dec.)\,$=$\,(\timeform{01h05m49s}, \timeform{-72D07'26''}), and the other (B in
236: figure~\ref{fg:f2}) is inside of the FoV at ($x$, $y$)\,$=$\,(8.5, $-$2) or (R.\,A.,
237: Dec.)\,$=$\,(\timeform{01h05m48s}, \timeform{-72D06'57''}). The former has the smaller
238: $\chi^{2}$ value of the fit. The simulated images and the best-fit profiles at the two
239: local minima are shown in figure~\ref{fg:f3} (a) and (b).
240:
241: The systematic uncertainty of this procedure was assessed using a different XIS
242: observation of 0836$+$714. A bright source is found in the image close to the edge but
243: well within the detector. We artificially truncated the image of the source and derived
244: the displacement of the real and reconstructed peak positions. The derived uncertainty
245: of $\sim$8\arcsec\ and $\sim$15\arcsec\ respectively for the $x$ and $y$ directions is
246: shown as a cross at the bottom left in figure~\ref{fg:f2}. We designate the source
247: Suzaku\,J0105--72.
248:
249: \begin{figure}
250: \begin{center}
251: \FigureFile(80mm,80mm){2.jpg}
252: \end{center}
253: \caption{
254: Close-up image of the transient source in gray scale. The observed peak position is
255: shown by a cross. The open circles are the positions of the nearby sources with labels
256: given in table~\ref{tb:t3}, while the open squares are the positions of the two faint
257: sources (C1 and C2; also in table~\ref{tb:t3}) found in the archived Chandra
258: images. The range of initial positions of the simulated source is given by the gray
259: line. The contours indicate 90\% and 99\% confidence ranges of the transient source
260: position. The systematic uncertainty of the position determination is indicated as a
261: cross at the bottom left.}\label{fg:f2}
262: \end{figure}
263:
264: \begin{figure*}
265: \begin{center}
266: \FigureFile(160mm,160mm){3.jpg}
267: \end{center}
268: \caption{
269: Simulated images and profiles at three different positions; (a) the local minimum A, (b) local minimum B, and (c) the second closest source in
270: table~\ref{tb:t3}. The upper panels show the simulated images in gray scale. The
271: intensity of the simulated and observed images in an arbitrary unit is shown with gray
272: and black contours, respectively. The assumed positions are shown with pluses. The
273: lower panels show the projected profiles of the simulated (lines) and the observed
274: (pluses) images.}\label{fg:f3}
275: \end{figure*}
276:
277: \subsection{Temporal Analysis}
278: We constructed a background-subtracted light curve of the transient source
279: (figure~\ref{fg:f4}). The source and background photons were accumulated from the solid
280: and dashed circles truncated by the field edge in figure~\ref{fg:f1}. The source circle
281: around the observed peak has a radius of 3\arcmin, which would include $\sim$90\% of the
282: photons from a point source if the whole region is inside the FoV. The background region
283: was selected from a region devoid of X-ray emission at the edge to have the same radius
284: and the same truncated fraction with the source region.
285:
286: We found a decline of the count rate in the light curve. We fitted the curve with a
287: constant count rate model, which was rejected by a $\chi^{2}$ test with a
288: $\chi^{2}/\rm{d.o.f}$ value of $\sim$93$/$10. No such trend was found in the background
289: light curve, which indicates that the decline is intrinsic to the transient source.
290:
291: We also searched for any periodicity of signals using the Lomb method \citep{lomb76} and
292: the epoch folding search technique. We found no significant periodicity, except for
293: those related to the Suzaku orbit.
294:
295: \begin{figure}
296: \begin{center}
297: \FigureFile(80mm,80mm){4.jpg}
298: \end{center}
299: \caption{
300: Background-subtracted source (thick pluses) and background (thin pluses) light curves
301: with a binning size of 5760~s, the period of the Suzaku orbit. Events taken with the
302: four XIS in 0.2--2.0~keV were merged. The origin of the time is the start time of the
303: observation.}\label{fg:f4}
304: \end{figure}
305:
306: \subsection{Spectral Analysis}
307: We constructed the background-subtracted spectrum in 0.4--2.0~keV for the FI chips and
308: 0.2--2.0~keV for the BI chip (figure~\ref{fg:f5}). The source and the background signals
309: were integrated from the same regions with the temporal analysis. The spectrum is very
310: soft with almost all photons below 2~keV and shows no conspicuous emission lines. In
311: order to fit the spectrum, we created the detector and mirror responses on the
312: observation date at two local minima (A and B in figure~\ref{fg:f2}) using the
313: \texttt{xisrmfgen} and \texttt{xissimarfgen} tools. The spectra and responses were
314: constructed independently for the four CCDs and the fitting trials were conducted
315: simultaneously for the set of four spectra. In order to account for the possible
316: normalization differences among the four CCDs, we added normalization parameters
317: relative to XIS1.
318:
319: We fitted the spectrum with several continuum (blackbody, bremsstrahlung, and power-law)
320: models with interstellar extinction. The blackbody model with a temperature of
321: \kt~$\sim$72~eV and the interstellar extinction of
322: \nh~$\sim$4.9$\times$10$^{20}$~H~atoms~cm$^{-2}$ yielded the best-fit result, which
323: explains the global spectral shape quite well. The amount of interstellar extinction is
324: consistent with the value toward the SMC \citep{dickey90}. However, we see an edge-like
325: residual at $\sim$0.74~keV as shown in the middle panel of figure~\ref{fg:f5}. We
326: therefore modified the blackbody model by adding an edge model with two more free
327: parameters, the edge energy ($E$) and the optical depth at the edge ($\tau$), and
328: obtained an improved acceptable fit. Based on the F-test, we found that the improvement
329: is statistically significant ($<$10$^{-6}$ of a chance probability).
330:
331: The best-fit parameters are given in table~\ref{tb:t2} separately for the position A and
332: B. X-ray flux (\fx) and luminosity (\lx) are derived in the 0.2--2.0~keV band. A
333: distance of 60~kpc is assumed for the luminosity. The best-fit model is given in
334: figure~\ref{fg:f5} for the position A. The parameters for the spectral shape (\nh, \kt,
335: $E$, and $\tau$) do not differ so much between the position A and B. This is because the
336: half power diameter and the vignetting function of the telescopes do not depend on
337: energy at $<$5~keV. On the other hand, the parameters for the normalization (\fx) are
338: different. This is because A is further away from the observed peak, thus requires a
339: brighter intrinsic flux.
340:
341: \begin{table*}
342: \begin{center}
343: \caption{Best-fit parameters of the fitting model}\label{tb:t2}
344: \begin{tabular}{lllll}
345: \hline\hline
346: Components & Par. & Units & Values (at pos.~A)\footnotemark[$*$] & Values (at pos.~B)\footnotemark[$*$] \\
347: \hline
348: Absorption & \nh & (cm$^{-2}$) & 4.87 $_{-0.42}^{+0.44}$~$\times$10$^{20}$ & 4.86 $_{-0.42}^{+0.44}$ $\times$10$^{20}$ \\
349: Blackbody & \kt & (eV) & 71.6 $_{-2.0}^{+2.1}$ & 71.6 $_{-2.0}^{+2.1}$ \\
350: & \fx\footnotemark[$\dagger$] & (erg~s$^{-1}$~cm$^{-2}$) & 1.02 $_{-0.06}^{+0.04}$~$\times$10$^{-11}$ & 4.66 $_{-0.45}^{+0.31}$ $\times$10$^{-12}$ \\
351: & \lx\footnotemark[$\dagger$] & (erg~s$^{-1}$) & 1.35 $_{-0.07}^{+0.05}$~$\times$10$^{37}$ & 6.14 $_{-0.30}^{+0.37}$ $\times$10$^{36}$ \\
352: Edge & $E$ & (keV) & 0.74 $_{-0.02}^{+0.02}$ & 0.74 $_{-0.02}^{+0.02}$ \\
353: & $\tau$ & & 1.22 $_{-0.39}^{+0.52}$ & 1.22 $_{-0.39}^{+0.43}$ \\
354: Constant & XIS0 & & 0.98 $_{-0.07}^{+0.07}$ & 1.00 $_{-0.07}^{+0.07}$ \\
355: & XIS1 & & 1.00 (fixed) & 1.00 (fixed) \\
356: & XIS2 & & 0.81 $_{-0.05}^{+0.06}$ & 0.94 $_{-0.06}^{+0.07}$ \\
357: & XIS3 & & 0.98 $_{-0.07}^{+0.08}$ & 1.05 $_{-0.08}^{+0.08}$ \\
358: \hline
359: \multicolumn{3}{l}{$\chi^{2}/\rm{d.o.f}$ ($\chi^{2}_{\rm{red}}$)} & 187.13/161 (1.16) & 187.29/161 (1.16) \\
360: \hline
361: \multicolumn{5}{@{}l@{}}{\hbox to 0pt{\parbox{130mm}{\footnotesize
362: \footnotemark[$*$] The uncertainties are for the 90\% confidence ranges, which
363: do not include systematic uncertainties in the instrumental calibration. See
364: http://www.astro.isas.jaxa.jp/suzaku/doc/suzaku\_td/ suzaku\_td.html for details.
365: \par\noindent
366: \footnotemark[$\dagger$] Flux and luminosity are estimated in a range of 0.2--2.0~keV.
367: The distance to the SMC (60~kpc) is assumed to derive the luminosity.
368: }\hss}}
369: \end{tabular}
370: \end{center}
371: \end{table*}
372:
373: \begin{figure}
374: \begin{center}
375: \FigureFile(80mm,80mm){5.jpg}
376: \end{center}
377: \caption{
378: XIS spectrum and the best-fit model. The background-subtracted spectra are shown with
379: pluses in the upper panel in different colors for the four censors (black, red, green,
380: and blue for XIS 0--3). The best-fit model (absorbed blackbody with an edge) is shown
381: with solid lines. The energies of the K edges are indicated by arrows. The lower two
382: panels show the residuals from the absorbed blackbody models with and without the
383: O\emissiontype{VII} edge.}\label{fg:f5}
384: \end{figure}
385:
386: \section{Discussion}
387: \subsection{Counterpart Search and Long-term Behavior}
388: First, in search for a possible bursts of this source in the past, we retrieved all the
389: available archived data taken in the vicinity by Einstein, ROSAT, ASCA, Beppo-SAX,
390: Chandra, and XMM-Newton. We inspected a total of 133 observation that span $\sim$28
391: years with an integrated exposure time of $\sim$2~Ms. No source was found in the error
392: region at a comparable brightness with the Suzaku source.
393:
394: Next, in search for the quiescent X-ray emission of this source and the counterpart in
395: other wavelengths, we further retrieved the SIMBAD database as well as the published
396: X-ray source lists using Einstein (\cite{inoue83}; \cite{seward81}; \cite{wang92}),
397: ROSAT (\cite{kahabka99a}; \cite{haberl00}; \cite{sasaki00}), and ASCA
398: \citep{yokogawa03}. Table~\ref{tb:t3} lists the sources within 2\arcmin\ of the observed
399: peak of the Suzaku source.
400:
401: We supplemented the list with two faint X-ray sources, which we found during the visual
402: inspection of the archived images. These two sources appear most significantly in a
403: series of Chandra images. We determined their positions, named them
404: CXOU\,J010540.1--720726 and J010550.3--720631, and refer to them as C1 and C2
405: hereafter. Four X-ray sources (sources 5, 6, C1, and C2) are listed in
406: table~\ref{tb:t3}. From the positional coincidence, we consider that the two ROSAT
407: sources (sources 5 and 6) and a Chandra source (C1) are the same.
408:
409: Among the sources in table~\ref{tb:t3}, source 1 is the only likely counterpart of the
410: Suzaku source for its positional coincidence including the systematic uncertainty. Other
411: sources including the two X-ray sources are displaced too far from both the islands A
412: and B, thus are unlikely to be the counterpart. At the position of source 2, the closest
413: source except for source 1, the simulated X-ray profile is inconsistent with the
414: observed one (figure~\ref{fg:f3}c).
415:
416: To reinforce our claim that the Suzaku source is different from the two nearby X-ray
417: sources, we reduced 26 data sets taken by the Advanced CCD Imaging Spectrometer (ACIS;
418: \cite{garmire03}) onboard Chandra \citep{weisskopf02} to reveal their nature. The
419: exposure times of all ACIS observations were too short to construct spectra. We derived
420: the photometric flux as the mean energy times the net count rate divided by the
421: accumulation area \citep{tsujimoto05} in the soft-band at each epoch and found that both
422: C1 and C2 are faint persistent sources (figure~\ref{fg:f6}). By combining all the data
423: sets, we constructed composite X-ray spectra (figure~\ref{fg:f7}). The two sources show
424: very similar spectra with hard and featureless emission. The absorbed power-law model
425: well explains these spectra. The best-fit index of power, 0.5--2.0~keV flux, and the
426: amount of extinction are 1.3--1.7, 4--5$\times$10$^{-15}$~erg~s$^{-1}$~cm$^{-2}$, and
427: 0.8--2.2$\times$10$^{21}$~cm$^{-2}$, respectively. The hard spectra extending beyond
428: $\sim$5~keV and the amount of extinction larger than that of the Suzaku source comprise
429: a sharp contrast with the Suzaku spectrum, indicating that C1 and C2 are not the Suzaku
430: counterpart. We examined a Digitalized Sky Survey (DSS) image at C1 and C2 and found no
431: optical source. From all these properties, both C1 and C2 are likely to be background
432: active galactic nuclei.
433:
434: Using the same Chandra data sets, we can derive the flux upper limits at the Suzaku
435: position for the last seven years and a half. Figure~\ref{fg:f6} shows a long-term trend
436: of the Suzaku and the two Chandra sources. The figure indicates that the transient
437: source experienced a burst during the Suzaku observation, and the flux was amplified by
438: more than 10$^{3}$ times compared to the undetected quiescent level.
439:
440: \begin{figure}
441: \begin{center}
442: \FigureFile(80mm,80mm){6.jpg}
443: \end{center}
444: \caption{
445: Long-term flux (0.5--2.0~keV) variation of the Suzaku and the neighboring Chandra
446: sources. The Suzaku transient flux at the positions A and B are plotted by filled
447: squares and circles, respectively. The 3~$\sigma$ upper limits at these positions by
448: Chandra are also plotted with open symbols. The flux of the two Chandra sources (C1 and
449: C2) are shown with upward and downward triangles, respectively. The typical uncertainty
450: for these sources are given for the first data points. Chandra data within ten days
451: were grouped.}\label{fg:f6}
452: \end{figure}
453:
454: \begin{figure}
455: \begin{center}
456: \FigureFile(80mm,80mm){7a.jpg}
457: \FigureFile(80mm,80mm){7b.jpg}
458: \end{center}
459: \caption{
460: ACIS spectra of the two Chandra sources (C1 and C2). The upper panels show the
461: background-subtracted spectra combining all the data sets (pluses) and the best-fit
462: absorbed power-law models (solid lines). The lower panels show the residuals to the
463: fit.}\label{fg:f7}
464: \end{figure}
465:
466: \begin{table*}
467: \begin{center}
468: \caption{List of nearby sources}\label{tb:t3}
469: \begin{tabular}{rlcccll}
470: \hline\hline
471: ID & Source name\footnotemark[$*$] & R.\,A. & Dec. & Separation\footnotemark[$\dagger$] & Object types\footnotemark[$*$] & References\footnotemark[$\ddagger$] \\
472: & & (J2000.0) & (J2000.0) & & & \\
473: \hline
474: 1 & 2dFS\,2064 & \timeform{01h05m46s} & \timeform{-72D06'51''} & \timeform{0.14'} & Star (B0) & [1] \\
475: 2 & MA93\,1554 & \timeform{01h05m50s} & \timeform{-72D06'33''} & \timeform{0.48'} & Emission-line star & [2] \\
476: 3 & FBR2002\,J010538--720643 & \timeform{01h05m38s} & \timeform{-72D06'43''} & \timeform{0.50'} & Radio source & [3] \\
477: 4 & 2dFS\,2059 & \timeform{01h05m44s} & \timeform{-72D07'28''} & \timeform{0.63'} & Star (F0) & [1] \\
478: 5 & RX\,J0105.7--7207 & \timeform{01h05m40s} & \timeform{-72D07'28''} & \timeform{0.70'} & X-ray source & [4] \\
479: 6 & HFP2000\,135 & \timeform{01h05m40s} & \timeform{-72D07'28''} & \timeform{0.70'} & X-ray source & [5] \\
480: 7 & MA93\,1542 & \timeform{01h05m39s} & \timeform{-72D06'09''} & \timeform{0.83'} & Emission-line star & [2] \\
481: 8 & OGLE\,SMC--SC10\,106764 & \timeform{01h05m36s} & \timeform{-72D06'22''} & \timeform{0.83'} & Star & [6] \\
482: 9 & OGLE\,SMC--SC10\,104288 & \timeform{01h05m35s} & \timeform{-72D07'22''} & \timeform{0.91'} & Unknown & [7] \\
483: 10 & MA93\,1548 & \timeform{01h05m43s} & \timeform{-72D08'11''} & \timeform{1.34'} & Emission-line star & [2] \\
484: 11 & 2dFS\,2011 & \timeform{01h05m24s} & \timeform{-72D06'44''} & \timeform{1.59'} & Star (B5) & [1] \\
485: 12 & MA93\,1531 & \timeform{01h05m27s} & \timeform{-72D07'45''} & \timeform{1.65'} & Emission-line star & [2] \\
486: 13 & 2dFS\,2110 & \timeform{01h06m06s} & \timeform{-72D06'20''} & \timeform{1.70'} & Emission-line star & [1] \\
487: 14 & IRAS\,F01038--7224 & \timeform{01h05m29s} & \timeform{-72D08'08''} & \timeform{1.73'} & Infra-red source & [8] \\
488: \hline
489: C1 & CXOU\,J010540.1--720726 & \timeform{01h05m40s} & \timeform{-72D07'26''} & \timeform{0.68'} & X-ray source & this work \\
490: C2 & CXOU\,J010550.3--720631 & \timeform{01h05m50s} & \timeform{-72D06'31''} & \timeform{0.55'} & X-ray source & this work \\
491: \hline
492: \multicolumn{7}{@{}l@{}}{\hbox to 0pt{\parbox{170mm}{\footnotesize
493: \footnotemark[$*$] The names and object types follow the SIMBAD designation except for C1 and C2.
494: The spectral types are given for stars in the object type when they are available.
495: \par\noindent
496: \footnotemark[$\dagger$] Separation from the observed peak.
497: \par\noindent
498: \footnotemark[$\ddagger$] [1] \citet{evans04}, [2] \citet{meyssonnier93}, [3] \citet{filipovi02}, [4] \citet{filipovi00},
499: [5] \citet{haberl00}, [6] \citet{udalski98}, [7] \citet{ngeow06}, [8] \citet{moshir90}.
500: }\hss}}
501: \end{tabular}
502: \end{center}
503: \end{table*}
504:
505: \subsection{Nature of Suzaku\,J0105--72}
506: The transient source has a very soft spectrum of a blackbody temperature of $\sim$72~eV
507: (figure~\ref{fg:f5}). Assuming a distance of 60~kpc, the bolometric luminosity is
508: $\sim$10$^{37}$~erg~s$^{-1}$ and the blackbody sphere has a radius of
509: $\sim$10$^{8}$~cm. The light curve (figure~\ref{fg:f4}) shows the gradual decline of the
510: X-ray flux. The long-term behavior (figure~\ref{fg:f6}) indicates that the source was in
511: a burst with the amplified flux of more than 10$^{3}$ times than the undetected
512: quiescent level.
513:
514: All these properties are commonly seen among supersoft sources (SSSs). SSSs are
515: considered to be a binary of a white dwarf and a companion star \citep{heuvel92} in the
516: SSS phase, in which mass accretion from the companion star fuels hydrogen burning on the
517: white dwarf surface. The X-ray spectra observed by CCD or proportional counter
518: spectrometers are fitted by a blackbody model of \kt~$\lesssim$~100~eV and the
519: bolometric luminosity of 10$^{36}$--10$^{38}$~erg~s$^{-1}$ \citep{greiner00}. We
520: therefore conclude that the nature of the transient source is a white dwarf binary,
521: which was in the declining end of the SSS phase during the Suzaku observation.
522:
523: SSSs show a variety of temporal behaviors. Some sources are constant supersoft X-ray
524: emitters fueled by continuous mass inflow and steady nuclear burning (e.g., CAL\,87;
525: \cite{heuvel92}) with occasional off-states (e.g, CAL\,83; \cite{kahabka98}). Others
526: have eruptive nature recognized as classical novae (e.g., RX\,J0513.9--6951;
527: \cite{reinsch99}), which can be recurrent (e.g., RS\,Oph; \cite{hachisu07}). The
528: transient nature of the Suzaku\,J0105--72 indicates that the source is in the latter
529: group.
530:
531: The possible optical counterpart, source 1, is a spectroscopically-identified B0
532: star. This source may be the companion star of the Suzaku source. We caution, however,
533: that many anonymous DSS sources can be found in the positional confidence range of the
534: Suzaku source. Follow-up studies are necessary to identify the companion star.
535:
536: \subsection{White Dwarf Mass}
537: Although the observational data are limited, we can obtain a rough estimate of the mass
538: of the white dwarf. Two independent estimates can be derived from the X-ray light curve
539: and the plasma temperature, both of which indicate that the mass is $\approx$1.2~$M_{\odot}$.
540:
541: First, the duration of the supersoft phase is a function of mass (figure~3 in
542: \cite{hachisu06}). The supersoft phase is seen at a later stage of bursts in classical
543: novae. In the optically thick wind model (\cite{hachisu06} and references therein), the
544: supersoft phase emerges when the wind stops and the inflated photosphere shrinks after
545: the optical burst and ends when the hydrogen shell burning is extinguished. As the mass
546: increases, the duration of the supersoft phase becomes shorter. This is because for a
547: larger mass, hence for a stronger surface gravity, the accreting mass can be smaller to
548: ignite the thermonuclear process, resulting in a shorter time scale to exhaust the fuel
549: \citep{hachisu05}.
550:
551: In the light curve of the supersoft burst (figure~\ref{fg:f6}), the Chandra observation
552: $\sim$200 days after the Suzaku burst gives a strong flux upper limit at the source
553: position. Besides, we do not find any significant emission at the position in the Suzaku
554: image obtained $\sim$110 days after the burst on 2005 December 16 (table~\ref{tb:t1}),
555: despite the facts that the image was exposed for three times longer and that the
556: observation has a roll angle to cover the source well within the XIS FoV. These suggest
557: that the supersoft phase faded out quickly in less than $\sim$3.5 month. We compared the
558: duration with the calculation for the smallest metallicity value (figure~3 and table~8
559: in \cite{hachisu06}) and found that the mass is estimated to be larger than
560: $\sim$1.2~$M_{\odot}$. The rapid decay is reminiscent to the recent supersoft burst from
561: the recurrent nova RS\,Oph \citep{osborne06}, which faded out in $\sim$60 days.
562:
563: Second, the effective temperature of the soft X-ray blackbody emission is also a
564: function of mass (figure~13 in \cite{ebisawa01}), in which the temperature becomes
565: larger as the mass increases. This is because for a larger mass, hence for a stronger
566: surface gravity, the plasma electron density increases, which suppresses the
567: ionization. To achieve the observed level of ionization, the plasma temperature has to
568: be higher \citep{ebisawa01}. We observed a color temperature of 72$\pm$2~eV, which we
569: assume is similar to the effective temperature. The mass is estimated to be
570: $\sim$1.0--1.2~$M_{\odot}$ for the plasma at a local thermal equilibrium.
571:
572: \subsection{Absorption Edge}
573: Some SSSs show absorption features in their blackbody spectra, which is thought to arise
574: from the absorption by the atmosphere of the white dwarf (\cite{heise94};
575: \cite{teesellng96}; \cite{hartmann97}). These features provide valuable tools to
576: diagnose the physical status of the atmospheric plasma.
577:
578: The Suzaku spectrum shows a conspicuous edge feature at $\sim$0.74~keV with an optical
579: depth of $\sim$1.2 (figure~\ref{fg:f5} and table~\ref{tb:t2}). We consider that this is
580: the O\emissiontype{VII} K edge at 0.7393~keV \citep{scofield01}. In contrast, we do not
581: see significant features of other K absorption edges by O\emissiontype{VIII} at
582: 0.8714~keV, N\emissiontype{VI} at 0.5521~keV, and N\emissiontype{VII} at 0.6671~keV
583: \citep{scofield01}. The 90\% confidence upper limits to the optical depth of these
584: features were measured to be $\sim$1.18, $\sim$0.06, and $\sim$0.25, respectively.
585:
586: The presence of the O\emissiontype{VII} K edge and the upper limit of the
587: O\emissiontype{VIII} K edge set an upper limit to the temperature of the
588: atmosphere. \citet{ebisawa01} calculated the fractional density of ions at various
589: ionization stages for several elements (C, N, O, Ne, and Fe), as a function of plasma
590: temperature (0--150~eV), and two electron densities ($n_{e}$\,$=$\,10$^{18}$ and
591: 10$^{19}$~cm$^{-3}$). The K edge absorption cross sections for O\emissiontype{VII} and
592: O\emissiontype{VIII} are $\sim$2.4 $\times$10$^{-19}$ and $\sim$1.2
593: $\times$10$^{-19}$~cm$^{2}$, respectively \citep{lang06}. The ratio of the optical depth
594: ($<$1.42) is converted to the ratio of the column density ($<$2.84) of
595: O\emissiontype{VIII} to O\emissiontype{VII}. From the ratio and figure~1 in
596: \citet{ebisawa01}, the atmospheric plasma temperature is constrained to be $\lesssim$58
597: and $\lesssim$68~eV for the electron density of 10$^{18}$ and 10$^{19}$~cm$^{-3}$,
598: respectively.
599:
600: The lack of N absorption edges does not necessarily indicate that the nitrogen is
601: deficient in terms of the N/O ratio from the solar abundance. The metallicity of the SMC
602: is deficient from the solar value by several times \citep{russell92}, which we canceled
603: by using the ratio N/O. From the upper limits of the N\emissiontype{VI} K edge, the
604: relative abundance of N/O is $\lesssim$0.46 assuming \kt\,$=$\,60~eV and
605: $n_{e}$\,$=$\,10$^{18}$~cm$^{-3}$, which does not contradict with the solar value of
606: $\sim$0.14 \citep{anders89}.
607:
608: Finally, we mention that the blackbody plus edge model may be an
609: oversimplification. Using a grating X-ray spectrometer, \citet{orio04} showed that the
610: spectrum of a SSS (CAL87) has in fact a complex of emission lines and absorption edges,
611: which appears as a blackbody spectrum in a study using a CCD spectrometer with a lower
612: resolution by $\sim$10~times \citep{ebisawa01}. Higher resolution spectroscopy
613: observations need be conducted in the future to confirm these results.
614:
615: \section{Summary}
616: Using the Suzaku XIS, we detected a transient source during a routine calibration
617: observation of 1E\,0102.2--7219. The source was detected at the edge of the XIS
618: image. We constrained its position to a $\sim$20\arcsec\ accuracy with the aid of
619: ray-tracing simulations and named it Suzaku\,J0105--72.
620:
621: No such transient source was found in the catalogued and archived data sets despite a
622: large number of observations, indicating that the burst event is quite rare. The source
623: showed a two-fold decline in flux during the Suzaku observation, probably at the
624: declining phase of a rare burst in which the flux was amplified by at least 10$^{3}$
625: times of the undetected quiescent level.
626:
627: The spectrum is explained by a blackbody emission with a temperature of $\sim$72~eV. The
628: measured extinction of $\sim$4.9$\times$10$^{20}$~cm$^{-2}$ is typical to the
629: SMC. Assuming a distance of 60~kpc, the source has a bolometric luminosity of
630: $\sim$10$^{37}$~erg~s$^{-1}$.
631:
632: From the X-ray temporal and spectral features, we concluded that Suzaku\,J0105--72 is a
633: SSS in the SMC. A short duration and a relatively high plasma temperature infer the
634: white dwarf mass of $\approx$1.2~$M_{\odot}$. The edge feature at $\sim$0.74~keV seen in
635: the XIS spectrum is probably due to O\emissiontype{VII} ions in the white dwarf
636: atmosphere. The presence of O\emissiontype{VII} K edge as well as the absence of
637: O\emissiontype{VIII}, indicate that the atmosphere has a plasma temperature of
638: $\lesssim$58--68~eV.
639:
640: \bigskip
641:
642: The authors thank the reviewer, Izumi Hachisu, for his insightful discussion. The Suzaku
643: data were obtained during the Science Working Group time. M.\,T. and M.\,M. are
644: financially supported by the Japan Society for the Promotion of Science. The Einstein,
645: ROSAT, ASCA, Beppo-SAX, Chandra, and XMM-Newton data were obtained through the High
646: Energy Astrophysics Science Archive Research Center Online Service provided by the
647: NASA/Goddard Space Flight Center. This research has made use of data obtained from Data
648: ARchives and Transmission System (DARTS), provided by PLAIN center at ISAS/JAXA, and of
649: the SIMBAD database, operated at CDS, Strasbourg, France.
650:
651:
652: \begin{thebibliography}{}
653: \bibitem[Anders \& Grevesse(1989)]{anders89} Anders, E., \& Grevesse, N.\ 1989, \gca,
654: 53, 197
655: \bibitem[Dickey \& Lockman(1990)]{dickey90} Dickey, J.~M., \& Lockman, F.~J.\ 1990,
656: \araa, 28, 215
657: \bibitem[Ebisawa et al.(2001)]{ebisawa01} Ebisawa, K., et al.\ 2001, \apj, 550, 1007
658: \bibitem[Evans et al.(2004)]{evans04} Evans, C.~J., Howarth, I.~D., Irwin, M.~J.,
659: Burnley, A.~W., \& Harries, T.~J.\ 2004, \mnras, 353,
660: 601
661: \bibitem[Filipovi{\'c} et al.(2000)]{filipovi00} Filipovi{\'c}, M.~D., Haberl, F.,
662: Pietsch, W., \& Morgan, D.~H.\ 2000, \aap, 353, 129
663: \bibitem[Filipovi{\'c} et al.(2002)]{filipovi02} Filipovi{\'c}, M.~D., Bohlsen, T.,
664: Reid, W., Staveley-Smith, L., Jones, P.~A., Nohejl,
665: K., \& Goldstein, G.\ 2002, \mnras, 335, 1085
666: \bibitem[Flanagan et al.(2004)]{flanagan04} Flanagan, K.~A., Canizares, C.~R., Dewey,
667: D., Houck, J.~C., Fredericks, A.~C., Schattenburg,
668: M.~L., Markert, T.~H., \& Davis, D.~S.\ 2004, \apj,
669: 605, 230
670: \bibitem[Fujimoto et al.(2007)]{fujimoto07} Fujimoto, R., et al.\ 2007, \pasj, 59, S133
671: \bibitem[Gaetz et al.(2000)]{gaetz00} Gaetz, T.~J., Butt, Y.~M., Edgar, R.~J., Eriksen,
672: K.~A., Plucinsky, P.~P., Schlegel, E.~M., \& Smith,
673: R.~K.\ 2000, \apj, 534, L47
674: \bibitem[Garmire et al.(2003)]{garmire03} Garmire, G.~P., Bautz, M.~W., Ford, P.~G.,
675: Nousek, J.~A., \& Ricker, G.~R., Jr.\ 2003, \procspie,
676: 4851, 28
677: \bibitem[Greiner(2000)]{greiner00} Greiner, J.\ 2000, New Astronomy, 5, 137
678: \bibitem[Haberl et al.(2000)]{haberl00} Haberl, F., Filipovi{\'c}, M.~D., Pietsch, W.,
679: \& Kahabka, P.\ 2000, \aaps, 142, 41
680: \bibitem[Hachisu \& Kato(2005)]{hachisu05} Hachisu, I., \& Kato, M.\ 2005, \apj, 631,
681: 1094
682: \bibitem[Hachisu \& Kato(2006)]{hachisu06} Hachisu, I., \& Kato, M.\ 2006, \apjs, 167,
683: 59
684: \bibitem[Hachisu et al.(2007)]{hachisu07} Hachisu, I., Kato, M., \& Luna, G.~J.~M.\
685: 2007, \apjl, 659, L153
686: \bibitem[Hartmann \& Heise(1997)]{hartmann97} Hartmann, H.~W., \& Heise, J.\ 1997, \aap,
687: 322, 591
688: \bibitem[Hayashi et al.(1994)]{hayashi94} Hayashi, I., Koyama, K., Ozaki, M., Miyata,
689: E., Tsumeni, H., Hughes, J.~P., \& Petre, R.\ 1994,
690: \pasj, 46, L121
691: \bibitem[Heise et al.(1994)]{heise94} Heise, J., van Teeseling, A., \& Kahabka, P.\
692: 1994, \aap, 288, L45
693: \bibitem[Hughes et al.(2000)]{hughes00} Hughes, J.~P., Rakowski, C.~E., \& Decourchelle,
694: A.\ 2000, \apj, 543, L61
695: \bibitem[Inoue et al.(1983)]{inoue83} Inoue, H., Koyama, K., \& Tanaka, Y.\ 1983, IAU
696: Symp.~101: Supernova Remnants and their X-ray
697: Emission, ed. J.~Danziger \& P.~Gorenstein, 101, 535
698: \bibitem[Ishisaki et al.(2007)]{ishisaki07} Ishisaki, Y., et al.\ 2007, \pasj, 59, S113
699: \bibitem[Kahabka(1998)]{kahabka98} Kahabka, P.\ 1998, \aap, 331, 328
700: \bibitem[Kahabka et al.(1999)]{kahabka99a} Kahabka, P., Pietsch, W., Filipovi{\'c} ,
701: M.~D., \& Haberl, F.\ 1999, \aaps, 136, 81
702: \bibitem[Kokubun et al.(2007)]{kokubun07} Kokubun, M., et al.\ 2007, \pasj, 59, S53
703: \bibitem[Koyama et al.(2007)]{koyama07} Koyama, K., et al.\ 2007, \pasj, 59, S23
704: \bibitem[Lang(2006)]{lang06} Lang,~K.~R.,\ 2006, in Astrophysical Formulae volume 1
705: (Springer-Verlag Berlin Heidelberg), 51
706: \bibitem[Lomb(1976)]{lomb76} Lomb, N.~R.\ 1976, \apss, 39, 447
707: \bibitem[Meyssonnier \& Azzopardi(1993)]{meyssonnier93} Meyssonnier, N., \& Azzopardi,
708: M.\ 1993, \aaps, 102, 451
709: \bibitem[Mitsuda et al.(2007)]{mitsuda07} Mitsuda, K., et al.\ 2007, \pasj, 59, S1
710: \bibitem[Moshir et al.(1990)]{moshir90} Moshir, M., et al.\ 1990, IRAS Faint Source
711: Catalogue, version 2.0
712: \bibitem[Ngeow \& Kanbur(2006)]{ngeow06} Ngeow, C.-C., \& Kanbur, S.~M.\ 2006, \mnras,
713: 369, 723
714: \bibitem[Orio et al.(2004)]{orio04} Orio, M., Ebisawa, K., Heise, J., \& Hartmann, J.\
715: 2004, Revista Mexicana de Astronomia y Astrofisica
716: Conference Series, 20, 210
717: \bibitem[Osborne et al.(2006)]{osborne06} Osborne, J., et al.\ 2006, The Astronomer's
718: Telegram, 838, 1
719: \bibitem[Rasmussen et al.(2001)]{rasmussen01} Rasmussen, A.~P., Behar, E., Kahn, S.~M.,
720: den Herder, J.~W., \& van der Heyden, K.\ 2001, \aap,
721: 365, L231
722: \bibitem[Reinsch et al.(1999)]{reinsch99} Reinsch, K., van Teeseling, A., Beuermann, K.,
723: \& Thomas, H.-C.\ 1999, Highlights in X-ray Astronomy,
724: 70
725: \bibitem[Russell \& Dopita(1992)]{russell92} Russell, S.~C., \& Dopita, M.~A.\ 1992,
726: \apj, 384, 508
727: \bibitem[Sasaki et al.(2000)]{sasaki00} Sasaki, M., Haberl, F., \& Pietsch, W.\ 2000,
728: \aaps, 147, 75
729: \bibitem[Sasaki et al.(2001)]{sasaki01} Sasaki, M., Stadlbauer, T.~F.~X., Haberl, F.,
730: Filipovi{\'c}, M.~D., \& Bennie, P.~J.\ 2001, \aap,
731: 365, L237
732: \bibitem[Sasaki et al.(2006)]{sasaki06} Sasaki, M., Gaetz, T.~J., Blair, W.~P., Edgar,
733: R.~J., Morse, J.~A., Plucinsky, P.~P., \& Smith,
734: R.~K.\ 2006, \apj, 642, 260
735: \bibitem[Scofield(2001)]{scofield01} Scofield,~J.~H.\ 2001, in X-ray Data Booklet,
736: ed. A.~Thompson et al. (Berkeley: University of
737: California), 1-8
738: \bibitem[Serlemitsos et al.(2007)]{serlemitsos07} Serlemitsos, P., et al.\ 2007, \pasj,
739: 59, S9
740: \bibitem[Seward \& Mitchell(1981)]{seward81} Seward, F.~D., \& Mitchell, M.\ 1981, \apj,
741: 243, 736
742: \bibitem[Takahashi et al.(2007)]{takahashi07} Takahashi, T., et al.\ 2007, \pasj, 59,
743: S35
744: \bibitem[Tsujimoto et al.(2005)]{tsujimoto05} Tsujimoto, M., Kobayashi, N., \& Tsuboi,
745: Y.\ 2005, \aj, 130, 2212
746: \bibitem[Udalski(1998)]{udalski98} Udalski, A.\ 1998, Acta Astronomica, 48, 113
747: \bibitem[van den Heuvel et al.(1992)]{heuvel92} van den Heuvel, E.~P.~J., Bhattacharya,
748: D., Nomoto, K., \& Rappaport, S.~A.\ 1992, \aap, 262,
749: 97
750: \bibitem[van Teesellng et al.(1996)]{teesellng96} van Teesellng, A., Heise, J., \&
751: Kahabka, P.\ 1996, IAU Symp.~165: Compact Stars in
752: Binaries, ed. J.~van~Paradijs et al., 165, 445
753: \bibitem[Wang \& Wu(1992)]{wang92} Wang, Q., \& Wu, X.\ 1992, \apjs, 78, 391
754: \bibitem[Weisskopf et al.(2002)]{weisskopf02} Weisskopf, M.~C., Brinkman, B., Canizares,
755: C., Garmire, G., Murray, S., \& Van Speybroeck, L.~P.\
756: 2002, \pasp, 114, 1
757: \bibitem[Yokogawa et al.(2003)]{yokogawa03} Yokogawa, J., Imanishi, K., Tsujimoto, M.,
758: Koyama, K., \& Nishiuchi, M.\ 2003, \pasj, 55, 161
759: \end{thebibliography}
760: \end{document}
761: