1: %\documentstyle[12pt,aaspp4]{article}
2:
3: %\documentstyle[emulateapj,psfig]{article}
4: \documentclass[12pt,preprint]{aastex}
5:
6:
7: \shorttitle{HARD X-RAY STECTRA OF CIR~X-1}
8: \shortauthors{Ding et al.}
9:
10:
11: \begin{document}
12:
13:
14: \title{EVOLUTION OF HARD X-RAY SPECTRA \\
15: ALONG THE BRANCHES IN CIR~X-1}
16:
17: \author{G. Q. Ding \altaffilmark{1,2}, J. L. Qu \altaffilmark{1}, and T. P. Li \altaffilmark{1,3}}
18: \affil{dinggq@mail.ihep.ac.cn, qujl@mail.ihep.ac.cn, litp@mail.ihep.ac.cn}
19:
20:
21: \altaffiltext{1}{Laboratory for Particle Astrophysics, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing, China}
22: \altaffiltext{2}{Graduate School of Chinese Academy of Sciences, Beijing, China}
23: \altaffiltext{3}{Department of Physics \& Center for Astrophysics, Tsinghua University, Beijing, China}
24:
25: \begin{abstract}
26: Using the data from the PCA and HEXTE on board the RXTE satellite, we investigate
27: the evolution of the 3-200 keV spectra of the peculiar low mass X-ray binary (LMXB)
28: Cir~X-1 along the branches on its hardness-intensity diagram (HID) from the vertical
29: horizontal branch (VHB), through the horizontal horizontal branch (HHB) and normal
30: branch (NB), to the flaring branch (FB). We detect a power-law hard component in the
31: spectra. It is found that the derived photon indices ($\Gamma$) of the power-law hard
32: component are correlated with the position on the HID. The power-law component dominates
33: the X-ray emission of Cir~X-1 in the energy band higher than $\sim 20$ keV. The fluxes
34: of the power-law component are compared with those of the bremsstrahlung component in
35: the spectra. A possible origin of the power-law hard component is discussed.
36: \end{abstract}
37: \keywords{binaries: general --- stars: individual (Circinus X-1) --- X-rays: binaries --- X-rays: general --- X-rays: individual (Circinus X-1)}
38:
39:
40: %\keywords{binaries: general---stars: individual (Circinus~X-1)---X-rays:
41: %binaries---X-rays: stars}
42:
43: \section{INTRODUCTION}
44: Circinus~X-1 (Cir~X-1), with an orbital period of 16.6 days and a distance of
45: 5.5 kpc (Case et al. 1998), is thought to be a peculiar X-ray binary.
46: Because its rapid variability is similar to that of Cyg~X-1, its
47: compact object was considered to be a black hole (Toor 1977). However, the compact
48: object is also believed to be a neutron star because of the presence of type I
49: X-ray bursts (Tennant et al. 1986 a, 1986 b). The radio counterpart of Cir~X-1,
50: coincident with a faint red star, displays flares with the same periodicity as
51: the orbital period (Whelan et al. 1977); its IR characteristics also
52: showed periodicity (Glass 1978); and its optical counterpart is proven to be a faint
53: red star (Moneti 1992). The soft X-ray (3-6 keV) flux of Cir~X-1 shows a well-defined
54: periodic modulation with the same period as its orbital period (Kaluzienski et
55: al. 1976). The large X-ray to optics luminosity ratio suggests that Cir~X-1 is
56: a LMXB.
57:
58: According to their fast timing properties and shapes of X-ray color-color
59: diagram (CCD) or the hardness-intensity diagram (HID), LMXBs whose primary compact
60: objects are neutron stars can be divided into two classes: Z source and atoll source
61: (Hasinger \& van der Klis 1989). There have been different opinions about what kind
62: of LMXB Cir~X-1 belongs to. From its temporal behaviors along its HID, Shirey(1998)
63: believed that Cir~X-1 is a Z source; but Oosterbroek et al.(1995), after having
64: studied its X-ray spectra and fast timing variations with EXOSAT ME data, suggested
65: that Cir~X-1 is the only known atoll source with a very low magnetic field strength
66: of less than $10^{9}$ G.
67:
68: The accretion geometry of LMXBs can be inferred from studies of their hard X-ray
69: spectra. Detailed studies of hard X-ray spectra of both Z and atoll sources have
70: been reported (Barret et al. 2000; Di Salvo et al. 2000; D'Amico et al. 2001; Di
71: Salvo et al. 2001). Using data of different high energy X-ray satellites
72: (Barret et al. 1999), the X-ray and hard X-ray spectra of LMXBs, including six Z
73: sources (Sco~X-1, GX~349+2, GX~340+0, Cyg~X-2, GX~5-1, and GX~17+2), have been
74: analyzed on the different branches of HID or CCD.
75:
76: Using the RXTE data, Shirey et al.(1999) investigated the continuous evolutions
77: of the Fourier power spectra and energy spectra of X-ray (2.5-25 keV) along
78: the branches of the HID of Cir~X-1, which displays a complete ``Z''
79: track. The quasi-periodic oscillation (QPO) frequencies from 12 Hz to 30 Hz have
80: been detected on the HB, 4 Hz on the NB, and no QPO on the FB. Also using RXTE
81: observation data, Qu et al.(2001) studied the evolution of time lags along the
82: complete HID; the time lag observed in Cir~X-1 seems to be consistent with the
83: Comptonization model with a two-layer corona.
84:
85: With the ASCA observations for Cir~X-1, Brandt et al.(1996) investigated the spectral
86: behavior for partial covering near zero phase of Cir~X-1, Iaria et al.(2001 a) reported
87: the spectral evolution along its orbit. Using the BeppoSAX data, Iaria et al.(2001 b, 2002)
88: analyzed the spectra of Cir~X-1 at the periastron and broadband spectrum at orbital phases
89: close to the apoastron. However, the high energy spectra of Cir~X-1 have been poorly
90: understood. In this work, using the PCA and HEXTE observations on board the RXTE, we
91: investigate the evolution of broadband spectrum (3-200 keV) of Cir~X-1 along its HID from
92: the VHB, through the HHB and NB, to the FB. The data selection and analysis are described
93: in \S 2, and results and discussion presented in \S 3.
94:
95: \section{OBSERVATIONS AND DATA ANALYSIS}
96:
97: We choose the Standard Mode 2 data of the PCA and Standard Modes (Archive) data of the
98: HEXTE to perform our analysis. Because detector 2 of cluster B of the HEXTE loses its
99: spectral capability and automatic gain control, only cluster A data are analyzed. The
100: used observations are listed in Table 1. The ftools Version5.2 is used to perform our
101: data analysis. The observations for which RXTE pointing offsets are larger than
102: $0.02\rm^o$ and elevation angles less than $10\rm^o$ are discarded.
103:
104: Using the method described by Shirey et al.(1999) with soft color being defined as
105: the ratio of the counts in the (6.5-13.0)/(2.0-6.5) keV energy bands, we use the PCA
106: data to obtain the HID of Cir~X-1, in which the soft color is plotted against the
107: intensity in the 2-18.6 keV energy interval. Figure 1 shows the obtained HID with
108: four branches sequentially called VHB, HHB, NB, and FB from the top left to the
109: bottom right of the diagram. In the VHB, the source intensities show little variation
110: with the soft color; in the HHB, the soft color remains almost constant with the source
111: intensity. The soft color is positively correlated with the source intensity in the
112: NB. The FB is connected to the soft end of the NB, where erratic variations are seen
113: in both the hardness ratio and the source intensity.
114:
115: We derive the hard X-ray spectrum of a certain branch according to the positions in
116: the HID for each observation. The resultant hard X-ray spectrum of a branch is then
117: obtained by adding all the spectra of the same branch from different OBSIDs. In order
118: to enhance the signal-to-noise ratio (SNR) of the HEXTE spectra, the energy channels
119: higher than 15 keV in the HEXTE spectra are rebinned. The energy channels are rebinned
120: in accordance with the principle: the higher the energy channels, the more the rebinned
121: energy channels.
122:
123: The spectra of 3-20 keV are derived from the PCA data (PCU0 only) and those of 20-200 keV
124: from the HEXTE data. The dips are discarded when the the X-ray spectra are extracted. The
125: HEXTE gaps and bad HEXTE observation intervals which are with zero photon count rate are
126: discarded when HEXTE spectra are extracted. We use Xspec 11.2 to analyze the four broadband
127: (3-200 keV) spectra. A systematic error of 1\% is added to each PCA spectrum due to the
128: calibration uncertainties. The PCA X-ray spectrum and HEXTE hard X-ray spectrum of every
129: branch are simultaneously fit by a combination model of a blackbody plus a line, a
130: bremsstrahlung, a power-law, and an absorption edge. The fit results are shown in Figures
131: 2 - 5 and Table 2. From the obtained spectra, we can estimate the absorbed fluxes of the
132: bremsstrahlung component of 3-200 keV, the power-law component of 20-200 keV, and the
133: luminosity of 3-200 keV. The results are listed in Table 3. The F-test (Protassov et al. 2000)
134: provided by Xspec 11.2 is performed to test the rationality of adding an extra power-law
135: component in the spectral fitting. For each branch, the fitting $\chi^2$ and d.o.f of the
136: Bblody+Line+Bremss+Edge model with and without an extra Power-law component, and the
137: corresponding F-test probability are listed in Table 4.
138:
139: \section{RESULTS AND DISCUSSION}
140:
141: From the F-test probability listed in Table 4, it is shown that adding an extra
142: power-law hard component is reasonable in the spectral fitting for each branch, which means
143: that the hard X-ray tail of Cir~X-1 is detected and its presence is not associated with a
144: particular position on its HID. The derived photon indices ($\Gamma$) of the four branches
145: span a range from 0.95 to 2.11 (see Table 2). The hardest index of the four branches is
146: obtained when the source is on the NB, and the softest is on the VHB.
147:
148: The mass accretion rate $\dot{M}$ increases in the sequence VHB $\rightarrow$ HHB
149: $\rightarrow$ NB $\rightarrow$ FB (van der Klis et al. 1996); the power law index
150: continuously decreases in the sequence VHB $\rightarrow$ HHB $\rightarrow$ NB
151: (see Table 2). In other words, the hard X-ray spectrum hardens from VHB, through HHB,
152: to NB with the increase of the mass accretion rate $\dot{M}$. So, there may exist a
153: correlation between the hardness of the hard X-ray spectrum and the mass accretion
154: rate $\dot{M}$. Our results show that the hard X-ray evolution of Cir~X-1 from the
155: horizontal branch (HB) to the NB is similar to that of Sco~X-1 from the HB to the
156: NB (D'Amico et al. 2001).
157:
158: From table 3, it is shown that hard X-ray luminosity fluxes fade in the sequence VHB
159: $\rightarrow$ HHB $\rightarrow$ NB; in other words, the less the mass accretion rate
160: $\dot{M}$, the larger the hard X-ray luminosity, and the softer the hard X-ray spectra.
161: This behavior of the hard X-ray spectra of Cir~X-1 is similar to that shown in Sco~X-1,
162: GX~17+2 (Di Salvo et al. 2000), and GX~349+2 (Di Salvo et al. 2001).
163:
164: A possible origin of the power-law hard component is the Compton up scattering of
165: soft X-ray seed photons by non-thermal electrons with a power-law velocity distribution
166: (Iaria et al. 2001 b). These electrons could be part of a jet in a binary system.
167: Similarly, a jet could also be the origin of the power-law hard component of the detected
168: hard X-ray spectra of some Z sources such as GX~5-1, GX~17+2 (Di Salvo et al. 2000), and
169: GX~349+2 (Di Salvo et al. 2001). So Cir~X-1 shares the same origin of the power-law hard
170: component with Z sources. It is another aspect that Cir~X-1 shares with the Z source family.
171:
172: The blackbody temperature is $\sim$ 1 keV, the blackbody component is probably emitted
173: by the inner part of the accretion disk (Iaria er al. 2002). The bremsstrahlung
174: seed-photon temperature of the Comptonized component is 1.2-3.1 keV. A Gaussian emission
175: line with energy $\sim$ 6.5 keV and an absorption edge with energy 8.2-9.2 keV are
176: detected. The iron ionization of Fe $_{\rm XXIV-XXV}$ corresponds to the emission line
177: which is compatible with the absorption edge (Turner et al. 1992).
178:
179: There shows an unusual line-like structure near 10 keV on the NB and FB (see Fig. 4 and Fig. 5).
180: This result is consistent with that reported by Shirey et al.(1999). Asai et al. discuss
181: several mechanisms that could possibly produce a line at $\sim$ 10 keV. For example,
182: emission from a heavy element such as Ni could produce the line; a line could be
183: blue-shifted due to motion in a relativistic jet or rotation in the accretion disk, but
184: extreme conditions would be required to boost the energy up to $\sim$ 10 keV
185: (Shirey et al. 1999). That there shows an unusual line near 10 keV in the FB spectrum
186: (see Fig. 5) is the reason why the reduced $\chi_{\nu}^2$ of this spectrum is as
187: high as 2.54.
188:
189: It is a key goal to distinguish neutron star (NS) binaries from black hole binaries
190: (BHBs) in high energy astrophysics. Barret et al.(2000) suggested that only BHBs can
191: have both X-ray (1-20 keV) and hard X-ray (20-200 keV) luminosities above
192: $1.5\times 10^{37}$ ergs s$^{-1}$. Using 5.5 kpc (Case et al. 1998) as the distance
193: to Cir~X-1, the luminosity of the power-law component in the 20-200 keV range is above
194: $1.5\times 10^{37}$ ergs s$^{-1}$. The reported total X-ray flux in the 2.5-25 keV band
195: of Cir~X-1 (Shirey et al. 1999) ranges over $(4.39-2.72)\times 10^{-8}$ ergs cm$^{-2}$
196: s$^{-1}$, thus the range of luminosity in the 2.5-25 keV range is
197: $L_{2.5-25}^{total}=(9.85-15.89)\times 10^{37}$ ergs s$^{-1}$, and the X-ray (1-20 keV)
198: luminosity must be above $1.5\times 10^{37}$ ergs s$^{-1}$. As is well known, the compact
199: object of Cir~X-1 has been believed to be a neutron star because of the presence of type
200: I X-ray bursts (Tennant et al, 1986 a, 1986 b), it seems that this luminosity criterion
201: used to distinguish NS binaries from BHBs is not suitable to Cir~X-1.
202:
203: \acknowledgments
204: We appreciate the detailed comments and suggestions provided by the anonymous
205: referee. This research has made use of data obtained through the High Energy
206: Astrophysics Archive Research Center On-line Service, provided by NASA/Goddard
207: Space Flight Center. We acknowledge the RXTE data teams at NASA/GSFC for their
208: help. This work is subsidized by the Special Funds for Major State Basic Research
209: Projects and by the National Science Foundation of China.
210:
211: \begin{thebibliography}{}
212: \bibitem[Brandt et al.(1996)]{bra96}Brandt, W. N., Fabian, A. C., Dotani, T.,
213: Nagase, F., Inoue, H., Kotani, T., \& Segawa, Y. 1996, \mnras, 283, 1071
214:
215: \bibitem[Barret et al. (2000)]{bar00} Barret, D., Olive, J. F., Boirin, L.,
216: Done, C., Skinner, G. K., \& Grindlay, J. E. 2000, \apj, 533, 329
217:
218: \bibitem[Case and Bhattacharya(1998)]{cas98} Case, G. L., \& Bhattacharya, D. 1998,
219: \apj, 504, 761
220:
221: \bibitem[D'Amico et al.(2001)]{d'am01} D'Amico, F., Heindl, W. A., Rothschild,
222: R. E., \& Gruber, D. E. 2001, \apj, 547, 147
223:
224: \bibitem[Di Salvo et al.(2000)]{dis00} Di Salvo, T., Stella, L., Robba, N. R.,
225: van der Klis, M., Burderi, L., Israel, G. L., Homan, J., Campana, S., Frontera,
226: F., \& Parmar, A. N. 2000, \apj, 544, 119
227:
228: \bibitem[Di Salvo et al.(2001)]{dis01} Di Salvo, T., Robba, N. R., Iaria, R.,
229: Stella, L., Burderi, L., \& Israel, G. L. 2001, \apj, 554, 49
230:
231: \bibitem[Glass(1978)]{gla78} Glass, I. S. 1978, \mnras, 183, 335
232:
233: \bibitem[Hasinger and van der Klis(1989)]{has89} Hasinger, G., \& van der Klis,
234: M. 1989, \aap, 225, 79
235:
236: \bibitem[Iaria et al.(2001 a)]{iar01a} Iaria, R., Di Salvo, T., Burderi, L., \&
237: Robba, N. R. 2001 a, ApJ, 561, 321
238:
239: \bibitem[Iaria et al.(2001 b)]{iar01b} Iaria, R., Burderi, L., Di Salvo, T., La
240: Barbera, A., \& Robba, N. R. 2001 b, \apj, 547, 412
241:
242: \bibitem[Iaria et al.(2002)]{iar02} Iaria, R., Di Salvo, T., Robba, N. R., \&
243: Burderi, L. 2002, \apj, 567, 503
244:
245: \bibitem[Kaluzienski et al.(1976]{kal76} Kaluzienski, L. J., Holt, S. S., Boldt,
246: E. A., \& Serlemitsos, P. J. 1976, \apj, 208, 71
247:
248: \bibitem[van der Klis et al.(1996)]{van96} van der Klis, M., Swank, J. H.,
249: Zhang, W., Jahoda, K., Morgan, E. H., Lewin, W. H. G., Vaughan, B., \& van
250: Paradijs, J. 1996, \apj, 469, 1
251:
252: \bibitem[Moneti(1992)]{mon92} Moneti, A. 1992, \aap, 260, 7
253:
254: \bibitem[Oosterbroek et al.(1995)]{oos95} Oosterbroek, T., van der Klis, M.,
255: Kuulkers, E., van Paradijs, J., \& Lewin, W. H. G. 1995, \aap, 297 141
256:
257: \bibitem[Protassov et al. (2000)]{pro00} Protassov, R. S., \& Dyk, D. A.
258: 2002, American Astronomical Society Meeting (AAS), 196, \#54.01
259:
260: \bibitem[Qu et al.(2001)]{qu01} Qu, J. L., Yu, W., \& Li, T. P. 2001, \apj,
261: 555, 7
262:
263: \bibitem[Rothschild et al.(1998)]{rot98} Rothschild, R. E., Blanco, P. R.,
264: Gruber, D. E., Heindl, W. A., MacDonald, D. R., Marsden, D. C., Pelling, M. R.,
265: Wayne, L. R., \& Hink, P. L. 1998, \apj, 496, 538
266:
267: \bibitem[Shirey(1998)]{shi98} Shirey, R. E. 1998, Ph.D. Thesis, MIT
268:
269: \bibitem[Shirey(1999)]{shi99} Shirey, R. E., Bradt, H. V., \& Levine, A. M.
270: 1999, \apj, 517, 472
271:
272: \bibitem[Toor(1977)]{too77} Toor, A. 1977, \apj, 215, 57
273:
274: \bibitem[Tennant et al.(1986 a)]{ten86a} Tennant, A. F., Fabian, A. C., \&
275: Shafer, R. A. 1986 a, \mnras, 219, 871
276:
277: \bibitem[Tennant et al.(1986 b)]{ten86b} Tennant, A. F., Fabian, A. C., \&
278: Shafer, R. A. 1986 b, \mnras, 221, 27
279:
280: \bibitem[Turner et al.(1992)]{tur} Turner, T. J., Done, C., Mushotzky, R., \&
281: Madejski, G. 1992, \apj, 391, 102
282:
283: \bibitem[Whelan et al.(1977)]{whe77} Whelan, J. A. J., Mayo, S. K., Wickramasinghe,
284: D. T., Murdin, P. G., Peterson, B. A., Hawarden, T. G., Longmore, A. J., Haynes,
285: R. F., Goss, W. M., Simons, L. W., Caswell, J. L., Little, A. G., \& McAdam, W.
286: B. 1977, \mnras, 181, 259
287: \end{thebibliography}
288:
289: \clearpage
290:
291: \begin{figure}
292: \vskip 2in
293: \special{psfile=f1.ps voffset=250 hoffset=2 vscale=67 hscale=55 angle=-90}
294: \vskip 1.85in
295: \caption{The HID of Cir~X-1. Each point is from 16 s background-subtracted data
296: from PCU012.}
297: \end{figure}
298:
299: \begin{figure}
300: \vskip 4.2in
301: \special{psfile=f2.ps voffset=250 hoffset=2 vscale=67 hscale=55 angle=-90 }
302: \vskip 1.85in
303: \caption{The VHB Spectrum of Cir~X-1. The spectrum results from 17570 s PCA (PCU0 only)
304: and 12290 s good HEXTE data. A combination model of Bbody+Line+Bremss+Power-law+Edge
305: fits to the observed spectrum with 1\% systematic error being set to the PCA
306: data. The reduced $\chi_{\nu}^2$ is 0.37.}
307: \end{figure}
308:
309: \begin{figure}
310: \vskip 2in
311: \special{psfile=f3.ps voffset=250 hoffset=2 vscale=67 hscale=55 angle=-90}
312: \vskip 1.85in
313: \caption{The HHB Spectrum of Cir~X-1. The spectrum results from 8130 s PCA (PCU0 only)
314: data and 12723 s good HEXTE data. A combination model of Bbody+Line+Bremss+Power-law+Edge
315: fits to the observed spectrum with 1\% systematic error being set to the PCA
316: data and the Line of E=6.50 keV being frozen. The reduced $\chi_{\nu}^2$ is 0.57.}
317: \end{figure}
318:
319: \begin{figure}
320: \vskip 2in
321: \special{psfile=f4.ps voffset=250 hoffset=2 vscale=67 hscale=55 angle=-90}
322: \vskip 1.85in
323: \caption{The NB Spectrum of Cir~X-1. The spectrum results from 10160 s PCA (PCU0 only)
324: data and 14980.0 s good HEXTE data and is fit by a combination model of Bbody+Line+Bremss+Power-law+Edge
325: with 1\% systematic error being set to the PCA data and the Line of E=6.50 keV being
326: frozen. The reduced $\chi_{\nu}^2$ is 0.94. There shows an unusual line near 10 keV
327: in this spectrum.}
328: \end{figure}
329:
330: \begin{figure}
331: \vskip 2in
332: \special{psfile=f5.ps voffset=250 hoffset=2 vscale=67 hscale=55 angle=-90}
333: \vskip 1.85in
334: \caption{ The FB Spectrum of Cir~X-1. The spectrum results from 6920 s PCA (PCU0 only)
335: data and 5130.0 s good HEXTE data and is fit by a combination model of Bbody+Line+Bremss+Power-law+Edge
336: with 1\% systematic error being set to the PCA data and the Line at 6.50 keV being frozen.
337: The reduced $\chi_{\nu}^2$ is 2.54. That there shows an unusual line near 10 keV in this
338: spectrum is the reason why the reduced $\chi_{\nu}^2$ of this spectrum is high.}
339: \end{figure}
340:
341: \clearpage
342:
343: \begin{table}
344: \caption{The used RXTE observations of Cir~x-1\label{tbl-1}}
345: \begin{tabular}{lcccc}
346: \tableline\tableline
347: & & & PCA observation & HEXTE observation \\
348: OBSID & MJD & HID Pos. & Time Interval (s) & Good Time Interval (s) \\
349: \tableline
350: 20094-01-02-020 & 49353 & VHB & 17570.0 & 12290.0 \\
351: 20094-01-02-020 & 49353 & HHB & 8130.0 & 6265.0 \\
352: 20094-01-02-020 & 49353 & NB & 2352.0 & 2299.0 \\
353: 20094-01-02-02 & 49353 & NB & 7808.0 & 6426.0 \\
354: 20094-01-02-02 & 49353 & FB & 6920.0 & 4290.0 \\
355: 20094-01-02-100 & 49353 & HHB & 12500.0 & 6458.0 \\
356: 20094-01-02-100 & 49353 & NB & 4672.0 & 2665.0 \\
357: 20094-01-02-09 & 49353 & FB & 1100.0 & 840.0 \\
358: 20094-01-02-09 & 49353 & NB & 11250.0 & 3590.0 \\
359: \tableline
360: \end{tabular}
361: \end{table}
362:
363: \begin{table}
364: \caption{The Fit Parameters \label{tbl-2}}
365: \begin{tabular}{lcccc}
366: \tableline\tableline
367: HID Pos. & VHB & HHB & NB & FB \\
368: \tableline
369: $ N_{H}$ $(\times 10^{22}$ cm$^{-2}$) & $3.02_{-0.60}^{+0.66}$ & $2.61_{-0.75}^{+1.56}$ & $12.88_{-2.86}^{+1.85}$ & $13.25_{-2.68}^{+2.46}$ \\
370: \tableline
371: $kT_{\rm bb}$(keV) & $1.021_{-0.059}^{+0.052}$ & $1.026_{-0.032}^{+0.023}$ & $0.95_{-0.14}^{+0.11}$ & $0.785_{-0.045}^{+0.049}$ \\
372: \tableline
373: $E_{\rm line}$(keV) & $6.48_{-0.43}^{+0.23}$ & 6.50 (frozen) & 6.50 (frozen) & 6.50 (frozen) \\
374: \tableline
375: $E_{\rm edge}$(keV) & $9.12_{-0.25}^{+0.34}$ & $9.19_{-0.19}^{+0.21}$ & $8.820_{-0.044}^{+0.085}$ & $8.284_{-0.068}^{+0.068}$ \\
376: \tableline
377: $kT_{\rm bremss}$(keV) & $2.92_{-0.50}^{+0.39}$ & $3.04_{-0.64}^{+0.46}$ & $2.89_{-0.44}^{+0.47}$ & $1.215_{-0.022}^{+0.030}$ \\
378: \tableline
379: $\Gamma_{\rm power-law}$(PhoIndex) & $2.11_{-0.69}^{+2.22} $ & $1.58_{-1.32}^{+8.42}$ & 0.95 (frozen) & $1.62_{-0.41}^{+0.69}$ \\
380: \tableline
381: $\chi^2~(dof)$ & 16.08 (44) & 23.36 (41) & 37.47 (40) & 81.21 (32) \\
382: \tableline
383: \end{tabular}
384: \tablecomments{Blackbody+Line+Bremsstrahlung+Power-law+Edge model is used. Uncertainties are given
385: at 90\% confidence level for the derived parameters of the model applied.}
386: \end{table}
387:
388: \clearpage
389:
390: \begin{table}
391: \caption{The Estimated Absorbed Fluxes and Luminosity \label{tbl-3}}
392: \begin{tabular}{lcccc}
393: \tableline\tableline
394: & Flux & Flux & Flux & Luminosity \\
395: & Bremsstrahlung & Power-law & All Components & All Components \\
396: HID Pos. & 3-200 keV & 20-200 keV & 3-200 keV & 3-200 keV \\
397: \tableline
398: VHB & 5.43 & 10.14 & 16.10 & 5.83 \\
399: HHB & 4.97 & 4.81 & 10.69 & 3.87 \\
400: NB & 2.81 & 0.40 & 5.36 & 1.94 \\
401: FB & 0.34 & 10.30 & 13.02 & 4.71 \\
402: \tableline
403: \end{tabular}
404: \tablecomments{The fluxes are in units of $10^{-8}$ ergs cm$^{-2}$ s$^{-1}$. The luminosity is in units
405: of $10^{38}$ ergs s$^{-1}$, assuming a distance of the source of 5.5 kpc (Case et al. 1988).}
406: \end{table}
407:
408: \begin{table}
409: \caption{F-test of The Four Spectra\label{tbl-4}}
410: \begin{tabular}{lccc}
411: \tableline\tableline
412: & Bbody+Line+Bremss & Bbody+Line+Bremss & \\
413: & +Edge & +Power-law+Edge & \\
414: HID Pos. & $\chi^2~ (dof)$ & $\chi^2~ (dof) $ & F-test Probability \\
415: \tableline
416: VHB & 31.48 (48) & 16.08 (44) & 4.49 $\times 10^{-6}$ \\
417: HHB & 31.64 (45) & 23.36 (41) & 1.27 $\times 10^{-2}$\\
418: NB & 60.99 (43) & 37.47 (40) & 1.94 $\times 10^{-4}$ \\
419: FB & 142.08 (36) & 81.21 (32) & 1.02 $\times 10^{-3}$ \\
420: \tableline
421: \end{tabular}
422: \end{table}
423:
424:
425:
426: \end{document}
427:
428:
429: