0705.4010/ms.tex
1: %% *************************************************************************
2: %%                                                               chjaa.tex
3: %% ChJAA Ver. 1.2, LaTeX class for Chinese Journal of Astronomy & Astrophysics
4: %% demonstration file
5: %%                                      (C) Chin. J. Astron. Astrophys.
6: %%                                      revised by Zhou Ai-Ying, 2001.08.28;2003.04.20;2006.09.07
7: %%
8: %% Note: 1. The ChJAA macro class chjaa.cls for LaTeX2e was originally adapted from
9: %%          aa.cls, the A&A's macro class (Ver.5.01). Manuscripts for
10: %%          ChJAA can be prepared following this demo file using chjaa.cls.
11: %%       2. To include EPS/PS graphics, you may use one of the graphics macro packages:
12: %%          graphicx.sty, epsf.sty or psfig.sty, or their mix. Please note the usage
13: %%          of \input{*.sty} and \usepackage{*}, they have no difference.
14: %%       3. Pay close attention to the format of ChJAA's reference list and
15: %%          other requirements. See Instructions for Authors (2001, ChJAA, Vol.1, No.1)
16: %%          or visit ChJAA's web pages at http://www.chjaa.org
17: 
18: 
19: %%       4. Final publication layout will be produced with this LaTeX source file.
20: %%---------------------------------------------------------------------------------
21: %%
22: \documentclass[referee]{chjaa}          % referee version: for submission
23: %\documentclass{chjaa}                  % preprint, the final version for publication
24:                                         %if use preprint, please de-comment \volnopage{} too.
25: 
26: \usepackage{graphicx,times}             %for PS/EPS graphics inclusion, new
27: \usepackage{lscape}
28: \input{epsf.sty}                        %for PS/EPS graphics inclusion, old
29: \input{psfig.sty}                       %for PS/EPS graphics inclusion, old
30: 
31: \headsep=6mm
32: \renewcommand{\baselinestretch}{1.15}
33: 
34: 
35: \begin{document}
36: 
37:    \title{Spectral study on the dips of Cir X-1
38: %\,$^*$
39: %\footnotetext{$*$ Supported by the National Natural Science Foundation of China.}
40: }
41: %   \subtitle{I. Place Your Subtitle Here}
42: 
43:    \volnopage{Vol.0 (200x) No.0, 000--000}      %%preserved for Editor. DOn't remove!
44:    \setcounter{page}{1}           %%starting page, preserved for Editor. DOn't remove!
45: 
46:    \author{Ya-Juan Lei
47:       \inst{1}\mailto{}
48: %% Please move "\mailto{}" to the corresponding author of the paper.
49: %% For single author or all the authors from an institute, use "\inst{}" only
50: %% Here is an example of three authors come from different institutes.
51:    \and Fang-Jun Lu
52:       \inst{1}
53:    \and Jin-Lu Qu
54:       \inst{1}
55:    \and Li-Ming Song
56:       \inst{1}
57:    \and Cheng-Min Zhang
58:       \inst{2}
59:       }
60: 
61:    \institute{Laboratory for Particle Astrophysics, Institute of
62:  High Energy Physics, Chinese Academy of Sciences, Beijing, 100049, China\\
63:              \email{leiyj@mail.ihep.ac.cn}
64: %% Please give the E-mail address of the author, to whom future correspondence and
65: %% offprint requests will be sent. Note to pair \mailto{} with \email{}
66:         \and
67:          National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100012, China   \\
68:         }
69: 
70:    \date{Received~~2007 month day; accepted~~2007~~month day}
71: 
72: \abstract{
73: We present X-ray spectral analyses of low mass X-ray binary Cir X-1 during X-ray dips, 
74: using the {\sl Rossi X-ray Timing Explorer (RXTE)} data.
75: Each dip was divided into several
76: segments, and the spectrum of each segment was fitted with a three-component blackbody
77: model, in which two components are affected by partial covering and the third one 
78: is unaffected. A Gaussian emission line is also included in the spectral model to 
79: represent the Fe K$\alpha$ line at $\sim$ 6.4\,keV. The fitted temperatures of  
80: the two partially covered 
81: components are about 2\,keV and 1\,keV, while the uncovered component
82: has a temperature of $\sim$0.5-0.6\,keV. The equivalent  
83: blackbody emission radius of the hottest component is the smallest and 
84: that of the coolest component is the biggest. During dips, the 
85: fluxes of the two hot components are linearly correlated, while that of the third
86: component doesn't show any significant variation.
87: The Fe line flux remains constant within errors during the short dips. However, 
88: during the long dips the line flux changes significantly and is positively 
89: correlated with the fluxes of the two hot components. These results suggest: 
90: (1) the temperature of the X-ray emitting region decreases with radius, (2) the 
91: Fe K$\alpha$ line emitting region is close to the hot continuum emitting region,
92: and (3) the size of the Fe line emitting region is bigger than the size of the 
93: obscuring matters causing short dips but smaller than the
94: sizes of those causing long dips.
95: \keywords{stars: individual (Circinus X-1) -- stars: neutron -- X-rays: stars}
96: }
97: 
98: 
99: 
100:    \authorrunning{Y.-J. Lei et al.}            %author_head in even pages
101:    \titlerunning{X-ray Spectroscopy of the dips of Cir X-1}  % title_head in odd pages
102: 
103:    \maketitle
104: %% The author head (on even pages) and the title head (on odd pages) will be
105: %% automatically extracted from \author{} and \title{}. Whenever the title is too long,
106: %% you will be asked to supply a shorter one by inserting either \authorrunning{} or
107: %% \titlerunning{} before \maketitle. Anyway, you can specify your own heads in advance.
108: %%
109: %%
110: %% Note: In the following text body of your manuscript, please note several differences from
111: %%       other major journals:
112: %% (1) \subsection{Please Capitalize the First Letter of Each Notional Word in Subsection Title}
113: %% (2) Please Capitalize the First Letter of Each Notional Word in all tables' captions
114: 
115: %
116: %________________________________________________ sections below
117: %
118: \section{Introduction}           %% first-level sections will be auto-capitalized
119: \label{sect:intro}
120: 
121: 
122: The X-ray light curves of some low mass X-ray binaries (LMXBs)
123: contain dips, which usually occur  near phase zero. In some
124: cases, the dips may be due to the disk instability resulting in the
125: rapid removal and replenishment of matter forming the inner part of
126: an optically thick accretion disk (e.g., Belloni et al. 1997;
127: Greiner et al. 1996). However, in most cases, the dips are suggested
128: to be due to the absorption by the  matter passing through the line
129: of sight to the X-ray emitting region (e.g., White et al. 1995).
130: The X-ray spectral evolution during the
131: dips can provide us information about the geometry  and
132: physical conditions of these regions in LMXBs (e.g., Asai et al.
133: 2000; Barnard et al. 2001).
134: 
135: The X-ray spectrum of a LMXB usually includes
136:  both the continuum and 6.4\,keV Fe K$\alpha$ line
137: emission components ({\sl e.g}., Hirano et al. 1987; White et al. 1986).
138: The continuum emission component is often described by the two
139: different models developed in the 1980's: the Western model
140: and   Eastern model. The ``Western model'' suggests that
141: Comptonization dominates the spectra, which has the form of a power
142: law with high energy cut off corresponding to the energy limit of
143: the Comptonizing electrons. However, for bright sources, an
144: additive  blackbody component representing the emission from the
145: neutron star (NS) is needed (e.g., White et al. 1988). The ``Eastern
146: model'' suggests that every observed spectrum contains two spectral
147: components: multi-temperature disk blackbody emission from the inner
148: disk and Comptonized emission provided by  the seed photons of NS
149: (e.g., Mitsuda et al. 1989). As declaimed, the physical conditions
150: of the emission region can be described by the ``Eastern model''.
151: For example, Done et al. (2002) assumed that there is an intrinsic
152: low energy cut off in the spectrum at $\sim$ 1\,keV due to
153: the lack of low energy seed photons, and the emission of NS  is
154: probably buried beneath an optically thick boundary layer.
155: As a result, the boundary layer dominates the hard  spectrum while
156: the disk dominates  the soft energies.
157: More recently, a model was proposed and developed by 
158: Church et al. (Church \& Ba\l uci\'nska-Church 1995; Church et al.
159: 1997; Ba\l uci\'nska-Church et al. 1999),  in which two continuum
160: components exist in the dipping LMXBs: the  simple blackbody
161: emission from NS plus Comptonized emission from an extended
162: accretion disk corona (ADC) above the accretion disk and a
163: ``progressive covering'' description of absorption. In addition,
164: this  model ascribes the dips to the absorption of obscuring matter
165: to the emission components.
166: For example, in X1755-338 and X1624-490, the absorption to the
167: blackbody component is primarily responsible for dipping (Church \&
168:  Ba\l uci\'nska-Church 1995, Church \&  Ba\l uci\'nska-Church
169: 1996), while in X1658-298 the absorption to the blackbody component
170: is lower than that to the Comptonized component (Oosterbroek et al.
171: 2001). In addition, for Cir X-1, Shirey et al. (1999) used a disk
172: blackbody  plus a blackbody and a ``progressive covering'' to fit the
173: spectra during the dips; Ding et al. (2006a) fitted the dip
174: spectra with the  two blackbodies plus a ``progressive covering'' for
175: a long dip of Cir X-1 (also see Brandt et al. 1996).
176: %(these two models are discussed in detail in section 2)}.
177: 
178: Fe K$\alpha$ line emission at 6.4 $\sim$ 6.7\,keV is a
179: common feature in the spectra of LMXBs, but its
180: origin is still in dispute. White et al. (1986) detected Fe K$\alpha$
181: emission line in five out of six LMXBs,
182: and proposed that the emission line is produced by the recombination
183: of Fe $XXVI$ in the inner ADC. Such a scenario was also suggested 
184: by Hirano et al. (1987). The Fe K$\alpha$ emission of dipping LMXBs 
185: is even more interesting since dips may provide new clues about
186: the location of the  line emitting region. 
187: Barnard et al. (2001) proposed that in XB 1323-619 the
188: Fe line at $\sim$ 6.4\,keV origins probably in the ADC,
189: while  Smale et al. (1993) ascribed the Fe line emission to the
190: reflection from the accretion disk itself for Cygnus X-2.
191: Using {\sl RXTE} data,  Shirey et al. (1999) investigated  the
192: spectral evolution of Cir X-1  during dips and found that the Fe
193: line flux is roughly constant during the dip transitions, suggesting
194: the Fe K$\alpha$ emission to be produced in the scattering medium.
195: 
196: Dips occur frequently on the light curve of Cir X-1, providing 
197: a good opportunity to study statistically the structure of the 
198: X-ray emitting regions and the location of Fe
199: K$\alpha$ line.
200: The spectral fits of the data of {\sl ASCA}, {\sl RXTE} and {\sl
201: BeppoSAX} during the dips are consistent with  a partial covering
202: model, in which the X-ray emission consists of a bright component
203: undergoing varying absorbing matter and a faint component unaffected
204: by the absorbing matter (e.g., Brandt et al 1996; Shirey et al. 1999; 
205: Iaria et al. 2001). The partial covering is a determinant of the behaviors
206: observed from Cir X-1, and matter
207: at the outer edge of the accretion disk, with an edge-on disk
208: orientation, could explain the partial covering of the spectrum (Brandt et al. 1996).
209: %{\bf 
210: Ding et al. (2006a) studied a long dip of Cir X-1 and concluded
211: that the covering matter exists from the surface of NS to the emitting region
212: of iron line in the accretion disk.
213: %}
214: However, another standpoints also exist, for example, D{\'{\i}}az Trigo et al. (2006) 
215: showed that there is no need for partial
216: covering to explain the spectral evolution during the dips with {\sl XMM} data.
217: 
218: In those previous studies of the dips of Cir X-1, each paper is usually based on 
219: the spectral analysis of a particular dip. We therefore present herein  a 
220: statistical study of the spectral evolution of Cir X-1 during various dips so as 
221: to better constrain the structure of the X-ray
222: emitting regions. The paper  is organized as follows: the observations, spectral models, 
223: and data analyses are described in Sect. 2, the results are shown in Sect. 3, 
224: the discussions are preseted in
225: Sect. 4 and the summary is in Sect. 5.
226: 
227: \section{Observations and Analysis}
228: \label{sect:Obs}
229: \subsection{Observations and Light Curves}
230: \label{subsect:data}
231: The Proportional Counter Array (PCA) of {\sl RXTE} consists of 5 non-imaging,
232: coaligned Xe multiwire proportional counter units (PCUs). It has a field of view
233: of 1$^{\circ}$ $\times$ 1$^{\circ}$ and a total
234: collecting area of $\sim$ 6500\,cm$^{2}$.
235: We use five datasets collected by PCU0 and PCU2 when both of them are working,
236: and both long (duration $\geq$ 10$^{4}$\,s) and short (duration $\leq$ 10$^{4}$\,s) 
237: dips are detected  in these data (Table 1). 
238: Since we are interested in the spectral evolution during dips, we
239: only study the  dips with durations $\geq$ 10$^{2}$\,s so
240: that each dip can be divided into a few segments and each segment
241: contains enough counts for spectral analysis.
242: Altogether we study the spectral properties of three long dips and
243: two short dips that are divided each into 5-17 segments.
244: Two typical light curves  with the denoted segments are shown in
245: Figs. 1 and 2.
246: 
247: The data are
248: reduced and analyzed using the {\sl Ftools} software package version
249: v5.2. We filter the data using the standard {\sl RXTE} criteria,
250: selecting time intervals for which parameters ${\tt Elevation\_Angles}
251: <10\degr$, ${\tt Time\_Since\_SAA} \geq 30$ min, ${\tt
252: Pointing\_Offsets} < 0.02 \degr$, and the background 
253: electron rate ${\tt Electron2} < 0.1$. 
254: Events in the {\sl RXTE} energy range $\sim$ 2.5-25\,keV are selected for
255: the spectral analysis. PCA background subtraction is carried out using the latest
256: versions of the appropriate background models, and a 1\% systematic error
257: is added to the spectra to account for calibration uncertainties.
258: As usual, the spectral fitting software $XSPEC$ is used.
259: 
260: \begin{table*}
261: %\begine{center}
262: \caption[]{Selected RXTE observations of Cir X--1\label{tbl-1}}
263: \begin{tabular}{l l l l}
264: \hline \hline
265:    OBSID              &  Start Time  &    Stop  Time   &     Exposure(s)\\
266: \hline
267:   10122-01-04-00      & 96-9-20 07:47:01 & 96-9-20 12:30:13 &    16992  \\
268: \hline
269:   10122-01-04-02      & 96-9-21 04:40:02 & 96-9-21 11:01:13 &    22871  \\
270: \hline
271:   30080-01-01-000     & 98-03-03 16:42:14 & 98-03-03 23:22:07 &    23993  \\
272: \hline
273:   30081-06-01-000     & 98-10-05 00:07:41 & 98-10-05 08:01:07 &    28406  \\
274:   30081-06-01-01      & 98-10-04 22:31:54 & 98-10-04 23:36:14 &    3860   \\
275: \hline
276:   60024-01-01-00      & 01-02-28 06:54:03 & 01-02-28 09:10:15 &    8172   \\
277:   60024-01-01-01      & 01-02-28 03:40:03 & 01-02-28 05:59:15 &    8352   \\
278:   60024-01-01-02      & 01-02-28 02:00:41 & 01-02-28 02:48:15 &    2854   \\
279:   60024-02-01-03      & 01-02-28 09:41:03 & 01-02-28 10:10:15 &    1752   \\
280: \hline
281: \end{tabular}
282: %\end{center}
283: \end{table*}
284: 
285: \begin{figure*}
286: \begin{center}
287: \includegraphics[width=6cm,angle=270,clip]{f1.ps}
288: \caption{Typical light curve (16\,s bins) of short ($<10^{4}\,s$) dip 
289: during dip and dip
290: transition from PCA observations of Cir X-1 (OBSID:30080). For clarity,
291: the errors are not plotted.}
292: \label{fig1}
293: \end{center}
294: \end{figure*}
295: 
296: %
297: \begin{figure}
298: \begin{center}
299: \includegraphics[width=6cm,angle=270,clip]{f2.ps}
300: \caption{Typical light curve (16\,s bins) of long ($>10^{4}\,s$) dip during dip and dip
301: transition from PCA observations of Cir X-1 (OBSID:30081).}
302: \label{fig2}
303: \end{center}
304: \end{figure}
305: 
306: 
307: \subsection{Spectral Model}
308: \label{subsect: model}
309: In the literature various spectral models have been used to fit the
310: spectra of Cir X-1 during dips (e.g., Brandt et al. 1996; Shirey et al. 1999;
311: Iaria et al. 2002; Ding et al. 2006a).
312: Brandt et al. (1996) used a model containing two blackbody
313: components and a partial covering. The partial covering can affect the hot
314: component, two or none of them.
315: Shirey et al. (1999) used a model which can be expressed as:
316: $F=[exp(-\sigma_{ph}N_{H}^{(1)})exp(-\sigma_{Th}N_{H}^{(1)})+exp(-\sigma_{ph}N_{H}^{(2)})f]M$,
317: where $F$ is the observed flux, $\sigma_{ph}$ and $\sigma_{Th}$ are
318: the photoelectric and Thomson cross sections, $N_{H}^{(1)}$ and
319: $N_{H}^{(2)}$ are the effective absorption hydrogen column densities to the
320: bright and faint components, respectively, $f$ is the ratio of the
321: unabsorbed flux of the faint component to the unabsorbed flux of the
322: bright component, and $M$ is the disk blackbody plus blackbody
323: model. Ding et al. (2006a) fitted the dip spectra with 
324: the ``Eastern'' model and a modified ``Western''
325: model. 
326: 
327: However, we find that all these above mentioned spectral models 
328: are improper to reveal the physical conditions of the dips. 
329: In Brandt et al. (1996) and Ding et al. (2006a), one of
330:  the fitted temperatures changes greatly within a dip. For an example, 
331: the temperature of the disk-blackbody  component
332: changes suddenly from $\sim$0.6\,keV to $\sim$1.2\,keV when its detected
333: fluxes evolves from the dip bottom to the transition state.  Such 
334: temperature changes seem unlikely if the dips are mainly due to the 
335: obscuration as widely accepted, because the obscuring matter can not
336: change the physical state of the emission region. 
337: In the model of Shirey et al. (1999), the residual flux at the bottom
338: of the dips is a fraction of the total emission. However, the 
339: hardness-intensity diagram (Fig. 3 in their paper and this paper)
340: shows that the detected spectrum of Cir X-1 at the lowest flux level is
341: even softer than that of its out-of-dip. This means that the residual 
342: emission spectrum at dip bottom is intrinsically softer  than that
343: nondip spectrum, i.e., it is improper to assume that
344: the residual emission is simply a small fraction of the total 
345: emission.
346: 
347: The soft spectrum at the bottom of the dip and the sudden temperature 
348: increase of the disk-blackbody component (Ding et al. 2006a) when Cir 
349: X-1 moves to the transition state, imply that there exists a weak and
350: soft component which is not affected by the obscuring matter. When
351: there is no dip or in the transition state, the usually called 
352: ``NS blackbody emission and disk-blackbody emission'' dominate, and so 
353: the weak component is difficult to be noted. When most of the ``NS blackbody emission 
354: and disk-blackbody emission'' are obscured when Cir X-1 is at the 
355: bottom of the dips, the contribution of the weak and soft component
356: to the total flux becomes significant or even dominant, and so the
357: detected spectrum at the dip bottom is softer than the
358: out-of-dip spectrum. Such a weak and soft component
359: is also noted by Ding et al. (2006a) but they do not consider it further
360: in the spectral fitting process.
361: 
362: Taking the above facts into account, we use in this paper a spectral
363: model containing three blackbody components and a Gaussian at 
364: the energy of Fe K$\alpha$ line. The 
365: first two blackbody components with partial covering are similar to those models 
366: in Brandt et al. (1996),  Shirey et al. (1999) and Ding et al. (2006a), 
367: while the additional third component represents the weak and soft emission that are
368: free of partial covering absorption. The expression of our model is: 
369: $F=wabs[pcfabs*cabs(bbodyrad(1)+bbodyrad(2)+gaussian)+bbodyrad(3)]$,
370: where $wabs$ represents the photoelectric absorption of the
371: interstellar medium (ISM), $pcfabs$ corresponds to the partial
372: covering photoelectric absorption, $cabs$ accounts for the Compton
373: scattering of the partial covering matter,
374: $bbodyrad(1)$ and $bbodyrad(2)$ represent the
375: continuum emission which is affected by the partial covering, $bbodyrad(3)$
376: is responsible for the soft faint emission component, and $gaussian$ accounts
377: for the Fe K$\alpha$ line emission.
378: As shown later, the fitted parameters by this model are
379: much more reasonable than those in Brandt et al. (1996), Shirey et al. (1999) 
380: and Ding et al. (2006a).
381: 
382: In the spectral fitting, the column density of $wabs$ is fixed as
383: 1.8$ \times 10^{22}$\,cm$^{-2}$ as determined in the
384: previous work (e.g., Brandt et al. 1996; Iaria et al. 2001), and the
385: column densities of $pcfabs$ and $cabs$ are forced to be identical (Shirey et al. 1999). 
386: As shown by the following fitting results (Figs. 4-9, Tables 2-3), such a model gives 
387: statistically acceptable fits to both the dip and nondip spectra,  and the
388: obtained parameters look physically reasonable.
389: 
390: %}
391: %
392: \begin{figure}
393: \begin{center}
394: \includegraphics[width=6cm,angle=270,clip]{f3.ps}
395: \caption{Hardness-intensity diagram for data in Fig. 2, intensity is I(2.0-18\,keV),
396: the hardness ratio is defined as : I(6.3-18\,keV)/I(2.0-6.3\,keV). 
397: Each point represents 16\,s of
398: background-subtracted data from all five PCA detectors (also see, Shirey et al. 1999).}
399: \label{fig3}
400: \end{center}
401: \end{figure}
402: 
403: %
404: \begin{figure}
405: \begin{center}
406: \includegraphics[width=6cm,angle=270,clip]{f4a.ps}
407: \includegraphics[width=6cm,angle=270,clip]{f4b.ps}
408: \caption{Top: Typical spectral fit for the dip spectra (spectrum C of OBSID 30080 in Table 2).
409: Bottom: Typical spectral fit for the nondip spectra (nondip data of OBSID 30080),
410: with $\chi^2$ values of 55.4 for 54 degrees of freedom.}
411: \label{fig4}
412: \end{center}
413: \end{figure}
414: 
415: 
416: \section{Result}
417: \label{sect:Result}
418: The spectra of all the segments have been fitted with the model
419: described in Sect. 2. Tables 2 and 3 list the
420: spectral fitting results for segments in OBSIDs 30080 and
421: 30081. In the following we will present our results in various aspects.
422: 
423: \begin{table*}
424: \caption[]{The Fitted Parameters for Spectra from the OBSID 30080 dip}
425: %\scriptsize{
426: \begin{tabular}{lccccc}
427: \hline \hline
428: Fit Parameters & A & B & C & D & E  \\
429: \hline
430: $N_{H}^{a}$ &
431:            1.8(fixed)              &     1.8(fixed)       &
432:        1.8(fixed)              &     1.8(fixed)       &
433:        1.8(fixed)               \\
434: \hline
435: $N_{H}^{b}$ &
436:     $0.0_{-0.0}^{+6.3}$     & $2.4_{-0.3}^{+0.6}$  &
437:     $16_{-8}^{+3}$          & $34_{-5}^{+6}$       &
438:     $38_{-21}^{+13}$           \\
439: \hline
440: $Percentage^c$ &
441:     $4.85_{-4.85}^{+95.15}$      &   95.2(fixed)            &
442:     $75.0_{-4.0}^{+24.4}$      & $89.9_{-2.9}^{+2.8}$     &
443:     $81.0_{-5.1}^{+4.2}$        \\
444: \hline
445: $kT_{\rm 1}^d$ (keV) &
446:     $1.85_{-0.04}^{+0.04}$     & $1.87_{-0.05}^{+0.05}$   &
447:         $1.91_{-0.08}^{+0.07}$   &
448:     $1.97_{-0.08}^{+0.10}$     &
449:     $1.89_{-0.08}^{+0.08}$      \\
450: %\hline
451: $A_{\rm 1}^e$ &
452:     $107_{-16}^{+9}$              & $93.7_{-8.8}^{+22.3}$    &
453:     $80.1_{-22.7}^{+21.4}$        & $61.9_{-10.7}^{+23.3}$     &
454:     $81.7_{-23.2}^{+36.3}$        \\
455: \hline
456: $flux_{\rm 1}^m$ &
457:     $1.11_{-0.16}^{+0.22}$       & $0.85_{-0.15}^{+0.22}$    &
458:     $0.49_{-0.13}^{+0.21}$        & $0.21_{-10.7}^{+23.3}$     &
459:     $0.26_{-0.08}^{+0.12}$        \\
460: 
461: \hline
462: $E_{\rm line}^f$ (keV) &
463:            6.60(fixed)         & $6.56_{-0.28}^{+0.33}$     &
464:     $6.48_{-0.49}^{+0.47}$     &
465:            6.60(fixed)     &
466:     $6.64_{-0.26}^{+0.34}$              \\
467: \hline
468: $\sigma^g$ (keV) &
469:           0.5(fixed)           &      0.5(fixed)       &
470:     $0.58_{-0.58}^{+0.70}$     &
471:     $0.36_{-0.36}^{+0.21}$     &   $0.22_{-0.22}^{+0.40}$     \\
472: \hline
473: $flux^h$ &
474:         $12.93_{-8.71}^{+5.76}$     &
475:     $16.98_{-9.21}^{+4.95}$    &
476:     $22.33_{-14.39}^{+47.80}$     & $14.14_{-2.95}^{+8.10}$ &
477:     $14.65_{-4.29}^{+16.54}$               \\
478: \hline
479: $kT_{\rm 2}^i$ (keV) &
480:     $1.00_{-0.04}^{+0.03}$     &
481:         $1.03_{-0.04}^{+0.01}$     &
482:     $1.06_{-0.12}^{+0.05}$     &
483:     $1.12_{-0.05}^{+0.06}$     &
484:     $1.08_{-0.04}^{+0.06}$      \\
485: %\hline
486: $A_{\rm 2}^j$ &
487:     $ 3.36_{-0.33}^{+0.43}\times10^{3} $     & $ 2.74_{-0.30}^{+0.25}\times10^{3} $   &
488:     $ 2.19_{-0.41}^{+0.58}\times10^{3} $      & $ 1.69_{-0.21}^{+0.42}\times10^{3} $   &
489:     $ 1.65_{-0.81}^{+0.95}\times10^{3} $        \\
490: 
491: \hline
492: $flux_{\rm 2}^n$ &
493:     $3.92_{-0.33}^{+0.43} $        & $2.65_{-0.28}^{+0.48}$   &
494:     $1.35_{-0.51}^{+0.50}$        & $0.57_{-0.14}^{+0.17}$    &
495:     $0.57_{-0.28}^{+0.34}$             \\
496: \hline
497: $kT_{\rm 3}^k$ (keV) &
498:           0.50(fixed)    &        0.50(fixed)    &
499:           0.50(fixed)          & $0.49_{-0.02}^{+0.02}$   &
500:           0.50(fixed)       \\
501: %\hline
502: $A_{\rm 3}^l$ &
503:     $4.60_{-4.38}^{+8.38}\times10^{3}$        & $0.0_{-0.0}^{+2.93}\times10^{3}$   &
504:     $0.02_{-0.02}^{+6.47}\times10^{3}$        & $8.45_{-3.83}^{+7.08}\times10^{3}$    &
505:     $5.32_{-2.16}^{+3.87}\times10^{3}$             \\
506: \hline
507: $flux_{\rm 3}^p$ &
508:     $0.19_{-0.18}^{+0.35} $        & $0.0_{-0.0}^{+0.12}$   &
509:     $0.0_{-0.0}^{+0.28}$        & $0.31_{-0.14}^{+0.26}$    &
510:     $0.22_{-0.07}^{+0.13}$             \\
511: \hline
512: $\chi^2~(dof)^p$ &
513:     68.63(46)                  & 42.82(46)               &
514:     35.51(44)                  & 30.46(45)               &
515:     24.04(44)                   \\
516: \hline
517: $\chi_{\nu}^{2q}$ &
518:     1.49                       & 0.93                     &
519:     0.81                       & 0.68                     &
520:     0.55                       \\
521: \hline
522: \end{tabular}
523: \vspace{0.2cm}
524: \parbox{6.8in}
525: {\baselineskip 5pt {\sc Note~} Model of wabs[pcfabs*cabs(bbodyrad(1)+gaussian+bbodyrad(2))]+wabs*bbodyrad(3) is used.
526: Uncertainties are given at 90\% confidence level.\\}
527: 
528: $^a$Interstellar column density, in units of $10^{22}$\,cm$^{-2}$\\
529: $^b$Partial covering column density, in units of $10^{22}$\,cm$^{-2}$\\
530: $^c$Partial covering percentage\\
531: $^d$The temperature of blackbody(1)\\
532: $^e$The normalization of blackbody(1)\\
533: $^m$The flux of blackbody(1) in the 2.0-10.0\,keV energy range and in units of photons
534: cm$^{-2}$\,s$^{-1}$\\
535: $^f$Gaussian line energy\\
536: $^g$Gaussian width\\
537: $^h$Flux of the gaussian line, in units of $10^{-3}$ photons\,cm$^{-2}$\,s$^{-1}$\\
538: $^i$The temperature of blackbody(2)\\
539: $^j$The normalization of blackbody(2)\\
540: $^n$The flux of blackbody(2) in the 2.0-10.0\,keV energy range and in units of photons\,
541: cm$^{-2}$\,s$^{-1}$\\
542: $^k$The temperature of blackbody(3)\\
543: $^l$The normalization of blackbody(3)\\
544: $^p$The flux of blackbody(3) in the 2.0-10.0\,keV energy range and in units of photons\,cm$^{-2}$\,s$^{-1}$\\
545: $^p$$\chi^2~(dof)$: $\chi^2$: Chi-Squared; $dof$: degrees of freedom\\
546: $^q$$\chi_{\nu}^2$: reduced chi-squared: $\chi^2$/$dof$\\
547: \end{table*}
548: 
549: \begin{landscape}
550: \begin{table*}
551: \caption[]{The Fitted Parameters for Spectra of the OBSID 30081 dip}
552: %\scriptsize{
553: \begin{tabular}{lccccccc}
554: \hline \hline
555: Fit Parameters & A & B & C & D & E & F & G  \\
556: \hline
557: $ N_{H}^{a}$ &
558:        1.8(fixed)              &     1.8(fixed)       &
559:        1.8(fixed)              &     1.8(fixed)       &
560:        1.8(fixed)               &
561:        1.8(fixed)              &
562:        1.8(fixed)               \\
563: \hline
564: $ N_{H}^{b}$ &
565:     $43_{-4}^{+3}$             & $81_{-42}^{+29}$    &
566:     $325_{-53}^{+39}$          & $247_{-40}^{+29}$    &
567:     $44_{-8}^{+10}$             &
568:     $23_{-4}^{+5}$             &
569:     $13_{-8}^{+5}$             \\
570: \hline
571: $Percentage^c$ &
572:     $98.1_{-9.2}^{+1.9}$       & $82.9_{-7.1}^{+7.6}$     &
573:     $89.6_{-5.3}^{+2.8}$       & $90.8_{-5.4}^{+2.8}$     &
574:     $98.0_{-7.3}^{+2.0}$       &
575:          95.2(fixed)           &
576:     $55.1_{-5.5}^{+44.9}$                \\
577: \hline
578: $kT_{\rm 1}^d$ (keV) &
579:     $2.14_{-0.03}^{+0.04}$     & $2.11_{-0.07}^{+0.08}$   &
580:      2.20(fixed)               & $2.25_{-0.19}^{+0.42}$   &
581:     $2.11_{-0.05}^{+0.05}$     &
582:     $2.10_{-0.04}^{+0.04}$     &
583:     $2.05_{-0.04}^{+0.03}$     \\
584: %\hline
585: $A_{\rm 1}^e$ &
586:     $37.3_{-5.1}^{+5.0}$       & $31.0_{-13.9}^{+45.4}$    &
587:     $36.9_{-5.6}^{+8.7}$       & $33.5_{-24.8}^{+29.9}$     &
588:     $37.7_{-7.4}^{+9.7}$       &
589:     $41.6_{-5.8}^{+6.8}$       &
590:     $50.0_{-9.3}^{+4.2}$        \\
591: \hline
592: $flux_{\rm 1}^m$ &
593:     $0.11_{-0.02}^{+0.02}$       & $0.06_{-0.03}^{+0.09}$    &
594:     $0.00_{-0.00}^{+0.00}$       & $0.01_{-0.01}^{+0.01}$     &
595:     $0.10_{-0.02}^{+0.32}$       &
596:     $0.23_{-0.03}^{+0.04}$       &
597:     $0.47_{-0.09}^{+0.04}$        \\
598: \hline
599: $E_{\rm line}^f$ (keV) &
600:      6.60(fixed)               &   $6.63_{-0.13}^{+0.11}$   &
601:     $6.56_{-0.04}^{+0.06}$     &
602:     $6.55_{-0.05}^{+0.05}$     &
603:     $6.64_{-0.20}^{+0.32}$     &
604:     $6.65_{-0.24}^{+0.27}$     &
605:     $6.42_{-0.40}^{+0.32}$              \\
606: \hline
607: $\sigma^g$ (keV) &
608:      0.50(fixed)               &   $0.38_{-0.13}^{+0.18}$     &
609:     $0.20_{-0.20}^{+0.12}$     &
610:     $0.26_{-0.15}^{+0.09}$     & $0.45_{-0.31}^{+0.24}$   &
611:     $0.39_{-0.39}^{+0.28}$     &
612:     $0.54_{-0.43}^{+0.30}$          \\
613: \hline
614: $flux^h$ &
615:     $7.36_{-3.40}^{+2.88}$     & $8.17_{-2.60}^{+4.86}$     &
616:     $4.90_{-2.81}^{+2.54}$     &
617:     $5.64_{-2.77}^{+3.94}$     & $14.63_{-6.14}^{+3.74}$ &
618:     $17.21_{-7.34}^{+12.25}$   &
619:     $29.31_{-13.68}^{+16.96}$                \\
620: \hline
621: $kT_{\rm 2}^i$ (keV) &
622:     $1.19_{-0.03}^{+0.03}$     & $0.84_{-0.22}^{+0.13}$           &
623:     $1.27_{-0.07}^{+0.07}$     &
624:         $1.29_{-0.08}^{+0.13}$     &
625:     $1.21_{-0.05}^{+0.06}$     &
626:     $1.17_{-0.03}^{+0.03}$     &
627:     $1.12_{-0.03}^{+0.02}$     \\
628: %\hline
629: $A_{\rm 2}^j$ &
630:     $ 0.89_{-0.15}^{+0.19}\times10^{3}$         & $ 2.13_{-1.35}^{+4.86}\times10^{3} $   &
631:     $ 4.11_{-1.73}^{+3.31}\times10^{3}$        & $ 2.15_{-1.09}^{+1.39}\times10^{3} $   &
632:     $ 0.68_{-0.17}^{+0.50}\times10^{3} $        &
633:     $ 1.16_{-0.26}^{+0.30}\times10^{3} $        &
634:     $ 1.33_{-0.46}^{+0.34}\times10^{3} $        \\
635: \hline
636: $flux_{\rm 2}^n$ (keV) &
637:     $0.23_{-0.04}^{+0.05}$     & $0.12_{-0.07}^{+0.26}$          &
638:     $0.09_{-0.05}^{+0.07}$     &
639:         $0.09_{-0.04}^{+0.05}$     &
640:     $0.17_{-0.08}^{+0.13}$     &
641:     $0.61_{-0.16}^{+0.16}$     &
642:     $1.40_{-0.47}^{+0.35}$      \\
643: \hline
644: $kT_{\rm 3}^k$ (keV) &
645:     $0.58_{-0.02}^{+0.02}$     & $0.54_{-0.07}^{+0.07}$   &
646:     $0.53_{-0.02}^{+0.01}$     & $0.51_{-0.01}^{+0.03}$   &
647:     $0.54_{-0.02}^{+0.02}$     &
648:     $0.48_{-0.09}^{+0.11}$     &
649:           0.50(fixed)              \\
650: %\hline
651: $A_{\rm 3}^l$ &
652:     $3.61_{-0.62}^{+0.70}\times10^{3}$        & $4.11_{-1.57}^{+2.51}\times10^{3}$   &
653:     $4.71_{-0.71}^{+1.06}\times10^{3}$        & $5.33_{-0.54}^{+0.97}\times10^{3}$   &
654:     $4.32_{-0.91}^{+1.23}\times10^{3}$        &
655:     $6.75_{-4.44}^{+18.16}\times10^{3}$       &
656:     $0.0_{-0.0}^{+6.85}\times10^{3}$          \\
657: \hline
658: $flux_{\rm 3}^p$ &
659:     $0.35_{-0.06}^{+0.06}$        & $0.26_{-0.08}^{+0.15}$    &
660:     $0.28_{-0.04}^{+0.03}$        & $0.26_{-0.05}^{+0.05}$    &
661:     $0.28_{-0.06}^{+0.08}$        &
662:     $0.22_{-0.16}^{+0.60}$        &
663:     $0.0_{-0.0}^{+0.23}$          \\
664: \hline
665: %\hline
666: $\chi^2~(dof)^p$ &
667:     17.68(45)                  & 20.32(43)               &
668:     23.32(44)                  & 21.17(43)               &
669:     15.54(43)                  &
670:     18.28(44)                  &
671:     18.45(44)                   \\
672: \hline
673: $\chi_{\nu}^{2q}$ &
674:     0.39                       & 0.47                     &
675:     0.53                       & 0.49                     &
676:     0.36                       &
677:     0.42                       &
678:     0.42                        \\
679: \hline
680: 
681: \end{tabular}
682: \vspace{0.2cm}
683: \parbox{6.8in}
684: {\baselineskip 5pt {\sc Note~} Model of wabs[pcfabs*cabs(bbodyrad(1)+gaussian+bbodyrad(2))]+wabs*bbodyrad(3) is used. Uncertainties
685: are given at 90\% confidence level for the derived parameters of the model applied.\\}
686: 
687: $^a$Interstellar column density, in units of $10^{22}$\,cm$^{-2}$\\
688: $^b$Partial covering column density, in units of $10^{22}$\,cm$^{-2}$\\
689: $^c$Partial covering percentage\\
690: $^d$The temperature of blackbody(1)\\
691: $^e$The normalization of blackbody(1)\\
692: $^m$The flux of blackbody(1) in the 2.0-10.0\,keV energy range and and in units of photons\,
693: cm$^{-2}$\,s$^{-1}$\\
694: $^f$Gaussian line energy\\
695: $^g$Gaussian width\\
696: $^h$Flux of the gaussian line, in units of $10^{-3}$ photons\,cm$^{-2}$\,s$^{-1}$\\
697: $^i$The temperature of blackbody(2)\\
698: $^j$The normalization of blackbody(2)\\
699: $^n$The flux of blackbody(2) in the 2.0-10.0\,keV energy range and in units of photons\,
700: cm$^{-2}$\,s$^{-1}$\\
701: $^k$The temperature of blackbody(3)\\
702: $^l$The normalization of blackbody(3)\\
703: $^p$The flux of blackbody(3) in the 2.0-10.0\,keV energy range and in units of photons
704: \,cm$^{-2}$\,s$^{-1}$\\
705: $^p$$\chi^2~(dof)$: $\chi^2$: Chi-Squared; $dof$: degrees of freedom\\
706: $^q$$\chi_{\nu}^2$: reduced chi-squared: $\chi^2$/$dof$\\
707: \end{table*}
708: \end{landscape}
709: %
710: 
711: 
712: 
713: 
714: \subsection{Partial covering column density and the 
715: correlations with other  parameters}
716: \label{subsect: partial covering}
717: Brandt et al. (1996) used a partial covering model to fit the spectra
718: before and after the dip transition of Cir X-1, and found a strong Fe
719: K$\alpha$ edge in the low-state spectra,
720: which indicates that the obscuring matter has a very high column density,
721: i.e, $N_{\rm H}$ $>$ 10$^{24}$\,cm$^{-2}$. Shirey et al. (1999) obtained 
722: a  similar result. Our results show that the hydrogen column density of 
723: partial covering absorption is between 10$^{22}$\,cm$^{-2}$ and  
724: 3.3 $\times$ 10$^{24}$\,cm$^{-2}$ (e.g., Fig. 5).
725: The large variation of column density could be due to the partial
726: covering and/or the inhomogeneity of the covering matter. 
727: 
728: Figures 5 and 6 show relations of the normalizations of 
729: the Fe K$\alpha$, blackbody(1) and blackbody(2) versus the 
730: partial covering absorption column density. 
731: Obviously, the normalization of Fe K$\alpha$ line has a positive
732: correlation with $N_{\rm H}$, while the correlations are weak
733: for the normalizations of blackbody(1) and blackbody(2). Since in
734: the spectral fitting, the absorption column density and normalization
735: are usually coupled with each other, in order to check whether
736: the correlation between the Fe K$\alpha$ strength and the absorption 
737: column density is real or not, we derive the normalization ratios 
738: of the Fe K$\alpha$ normalization to the normalizations of blackbody(1)
739: and blackbody(2). As shown in Figs. 7 and 8, such coupling-free ratios
740: are also correlated with the absorption column density, indicating
741: that the origin of the Fe emission line is related to the
742: obscuration matter.
743: 
744: \begin{figure}
745: \begin{center}
746: \includegraphics[width=10cm,clip]{f5.ps}
747: \caption{The
748: normalization of Fe line is plotted against $N_{\rm H}$ (in units of
749: $10^{22}$\,cm$^{-2}$), and an obvious correlation can be found.}
750: \label{fig5}
751: \end{center}
752: \end{figure}
753: %
754: 
755: %
756: \begin{figure}
757: \begin{center}
758: \includegraphics[width=10cm,clip]{f6.ps}
759: \caption{The
760: normalizations of blackbody(1) (bottom) and blackbody(2) (top) are
761: plotted against $N_{\rm H}$, and both show the correlations.}
762: \label{fig6}
763: \end{center}
764: \end{figure}
765: %
766: 
767: %
768: \begin{figure}
769: \begin{center}
770: \includegraphics[width=10cm,clip]{f7.ps}
771: \caption{Ratio of the normalization of Fe line to that of
772: blackbody(1) is plotted against $N_{\rm H}$. The bottom panel is a
773: four-point binned plot of the top one, a positive
774: correlation can be seen.}
775: \label{fig7}
776: \end{center}
777: \end{figure}
778: 
779: %
780: \begin{figure}
781: \begin{center}
782: \includegraphics[width=10cm,clip]{f8.ps}
783: \caption{Similar to Fig. 7, but  blackbody(2)
784: is substituted.}
785: \label{fig8}
786: \end{center}
787: \end{figure}
788: %
789: 
790: 
791: 
792: \subsection{Temperatures and Sizes of emission regions}
793: \label{subsect: temperatures and sizes}
794: For every dip, the temperature of blackbody(1) is about 2\,keV and that
795: of blackbody(2) is about 1\,keV, which is consistent with the  results of
796: Shirey et al. (1999) with {\sl RXTE}. 
797: The temperature of blackbody(3), unaffected by the partial covering,
798: is about 0.5$\sim$0.6\,keV, which is consistent with
799: the result of Brandt et al. (1996) with {\sl ASCA}.
800: The almost constant values of the temperature also show that the selecting 
801: of a three-blackbody-component spectral model is proper. 
802: 
803: The distance to Cir X-1 is about 6-10\,kpc (e.g., Stewart et al. 1991; 
804: Goss \& Mebold 1977), so we adopt a value of 8 kpc in converting the blackbody
805: normalization  to radius.
806: Hence, the obtained radii of blackbody(1) (blackbody(2)) are
807: mostly in the range of 4-7\,km (16-30\,km).
808: As for the blackbody(3), usually, its radii lie in the range of
809: 30-50\,km. Since the typical radius of a NS is 10\,km, blackbody(1) may come 
810: from hot regions near/on the NS surface, while blackbody(2) and blackbody(3) 
811: are perhaps from the inner and outer disk respectively.
812: 
813: \subsection{Flux and Flux correlations}
814: \label{subsect: correlation}
815: In Fig. 9 we plot the relations between the fluxes of blackbody(1),
816: blackbody(2) and blackbody(3). A strong correlation between the 
817: fluxes of blackbody(1) and blackbody(2) is found, with a correlation
818: coefficient $0.98_{-0.04}^{+0.02}$. 
819: However, further investigations have not shown any sign of 
820: clear correlations between the flux of blackbody(3) and that of
821: blackbody(1) or blackbody(2).
822: 
823: We now show that the linear correlation between the fluxes
824: of blackbody(1) and blackbody(2) is intrinsic other than a false 
825: appearance of the spectral fitting. In the spectral fitting, 
826: blackbody(1) and 
827: blackbody(2) are all coupled with the partial covering. When the 
828: partial covering column 
829: density (and/or covering fraction) is somehow overestimated, 
830: the normalizations of the two components will be in turn higher than
831: the real values, and vice versa. Then the two normalizations
832: will show an artificial positive correlation. However, the fluxes plotted in
833: Fig. 9 are the absorbed ones (corresponding to the observed fluxes), 
834: in which the model coupling effect is eliminated. Therefore, the
835: linear correlations shown in Fig. 9 is reliable. 
836: 
837: The flux correlation between blackbody(1) and blackbody(2) implies
838: that these two components are usually obscured simultaneously. 
839: So their emission regions are probably very close to each other.
840: This is well consistent with the conclusions we drew from the equivalent
841: emitting radii in the previous section.
842: 
843: 
844: %
845: \begin{figure}
846: \begin{center}
847: \includegraphics[width=10cm,clip]{f9.ps}
848: \caption{The flux
849: (in units of photons\,cm$^{-2}$\,s$^{-1}$ ) of blackbody(1)
850: plotted against that of blackbody(2) (top panel) and blackbody(3)
851: (bottom panel). The top panel clearly shows a linear correlation,
852: whereas  no significant relationship has been revealed in the bottom
853: panel.}
854: \label{fig9}
855: \end{center}
856: \end{figure}
857: 
858: \subsection{Fe line flux and its correlations with continuum components}
859: \label{subsect: Fe line flux}
860: %{\bf 
861: %Since we are interested in the properties of the iron line flux,
862: %including its origin  and the correlations between its flux
863: % and continuum spectral components in the dips.
864: %}
865: Figure 10 plots the Fe line fluxes versus the fluxes of the 
866: continuum components. It is found that the flux of the Fe line 
867: has a trend to increase along with the fluxes of blackbody(1) 
868: and blackbody(2), while no clear correlation
869: exists between the Fe line flux and that of blackbody(3).
870: Since the flux changes for both the line and continuum 
871: emission are mainly due to the obscuration of partial covering matters,
872: this correlation suggests that the Fe line emission origins close
873: to the hot regions that emit blackbody(1) and blackbody(2). 
874: 
875: We further divide the dips into long and short dips, and for
876: each category, we investigate the above correlations. As plotted in
877: Fig. 11, for long dips, the flux of Fe line has now even better 
878: correlations with those of the blackbody(1)
879: and blackbody(2) than the correlations for all the dips. However,
880: for short dips, such correlations do not exist anymore (see Fig. 12),
881: which is consistent with the results of Shirey et al. (1999). 
882: Similar to that for all the dips, correlations between the flux of the Fe line
883: and that of blackbody(3) are not detected either for long dips or for
884: short dips. These results tell us
885:  that the emitting region of the Fe line is smaller than the the obscuration matters
886: of long dips but bigger than those of the short dips.
887: 
888: 
889: %For the short dips, P30080 (see Fig. 1) for instance, the Fe
890: %line flux has little change within the errors during  dipping, even
891: %at the lowest portion of the dip (see the column E of Table. 2),
892: %which is consistent with the conclusions of Shirey et al. (1999).
893: 
894: %For the long dips, P30081 (see Fig. 2) as an example, its time
895: %of dip plus dip transition  is about 20000 s.
896: %In the initial 8000 s, the lower the counts rate,
897: %the smaller the Fe line flux (4.8$\sim$6.2 $\times$ 10$^{-3}$
898: %photons cm$^{-2}$ s$^{-1}$).
899: %However, at the dip transition time from 17300 to
900: %20000 s, the Fe line flux is almost unchanged
901: %($\sim$ 20 $\times$ 10$^{-3}$
902: %photons cm$^{-2}$ s$^{-1}$), which is obviously bigger  than
903: %that of the initial 8000 s (see the row $flux^{h}$ of Table.3 for P30081).
904: %The same results are also obtained from the other long dips.
905: %For all the long dips, the Fe line flux changes with the counts
906: %rate. Sometimes, at the peak of light curve, the best fitting line
907: %normalization is zero, which could be due to the fact that the flux of iron
908: %emission line is too faint,  comparing to the fluxes of continuum
909: %spectra,  to be constrained well.
910: %In this situation, the upper limit of iron line flux is presented.
911: 
912: 
913: 
914: %{\bf As indicated by Shirey et al. (1999) who study the dips of Cir X-1,
915: %the iron emission line is constant and associated with the faint component
916: %and might come from the scattering medium. Our results show that the iron
917: %emission line is not associated with the soft component which is unaffected
918: %by partial covering, but is correlated to the high temperature
919: %blackbody, might come from the inner emission region.}
920: %
921: 
922: %
923: \begin{figure*}
924: \begin{center}
925: \includegraphics[width=13cm,clip]{f10.ps}
926: \caption{Fe line
927: flux (in units of 10$^{-3}$ photons\,cm$^{-2}$\,s$^{-1}$)
928: plotted as a function of the flux of  blackbody(1) (left),
929: blackbody(2) (middle) and blackbody(3) (right),  for  all chosen
930: dips (10122-dip1, 30081, 60024, 10122-dip2, 30080).}
931: \label{fig10}
932: \end{center}
933: \end{figure*}
934: %
935: 
936: 
937: 
938: \begin{figure*}
939: \begin{center}
940: \includegraphics[width=13cm,clip]{f11.ps}
941: \caption{Similar to  Fig. 10, but for the
942: long dips (10122-dip1, 30081, 60024). The left (middle)
943: panel shows a correlation, and no significant
944: relationship has been noticed in the right one.}
945: \label{fig11}
946: \end{center}
947: \end{figure*}
948: %
949: 
950: \begin{figure*}
951: \begin{center}
952: \includegraphics[width=13cm,clip]{f12.ps}
953: \caption{Similar to  Fig. 10, but for the
954: short dips (10122-dip2, 30080).  The Fe line flux is almost
955: unchanged, within the errors, regarding to all continuum spectrum
956: fluxes.}
957: \label{fig12}
958: \end{center}
959: \end{figure*}
960: 
961: 
962: 
963: %
964: \section{Discussion}
965: \label{sect:discussion}
966: 
967: The binary system of Cir X-1 has a highly eccentric orbit. Possibly,
968: the NS may be occulted by the outer layers of its companion
969: atmosphere when the star moves close to the periastron.
970: The high eccentricity of the orbit might produce large tidal
971: interactions at this phase, in turn, which perhaps expand and/or
972: deform the  accretion stream, or maybe form a bulge in the outer accretion disk.
973: The excess of these matters may influence on the observed emission
974: of X-ray source, producing a partial covering of the energy
975: spectrum of the source (e.g., Inoue 1989; Iaria et al. 2001).
976: 
977: The Fe K$\alpha$ line emission at 6.4$\sim$6.7\,keV is a
978: common feature in LMXBs, and its origin has been an interesting
979: topic. Generally, there are four possible sites  for the Fe
980: K$\alpha$ emissions: the accretion disk, ADC, the source itself, or
981: the scattering medium (e.g., Smale et al. 1993; Shirey et al. 1999; Asai et al.2000).
982: The observed strong broad line at 6.4$\sim$6.7\,keV is
983: generally interpreted as arising from the ADC and illuminated outer
984: disk (e.g., White et al. 1985; Hirano et al. 1987; Asai et al.
985: 2000), but it is difficult to obtain both the observed line
986: intensity and width from such a corona.
987: 
988: %{\bf  
989: For Cir X-1, the results of Shirey et al. (1999) show that
990: the Fe K$\alpha$ line flux is constant and associated with the faint 
991: component during short dips,
992: implying that the Fe K$\alpha$ comes from the scattering medium far
993: away from the central region.
994: After investigating  the long dip of Cir X-1, Ding et al. (2006a,b)
995: propose that Fe line emission is from the disk.
996: Our results show that the flux of iron line emission changes
997: obviously during the long dips (see OBSID 30081 of Table 3) and is positively
998: correlated with the fluxes of  blackbody(1) and blackbody(2).
999: This token implies that the line could not be associated with the
1000: soft component unaffected by the partial covering,
1001: and instead, it is affected by the partial covering and could come 
1002: from the region close to the central object.
1003: 
1004: %Our result is consistent with that of Ding et al. (2006) which shows that 
1005: %the covering matter is not distributed over the total disk, but extended to 
1006: %the region of iron line emission.
1007: 
1008: For short dips, our result shows that the flux of the Fe
1009: emission line is almost constant.
1010: There are two possible interpretations for this constant Fe line emission
1011: flux: (1) for the short dip the obscuring
1012: matter could be of a small size, and (2) the obscuring matter lies inner 
1013: than the Fe line emission materials on the accretion disk. 
1014: On the other hand, according to the fact that the normalization of
1015: the Fe line is positively correlated with $N_{\rm H}$ (see Fig. 5),
1016: we also considered that the Fe line could come from the accreting
1017: matter, which suggests that the Fe line strength is related to
1018: mass accretion rate. 
1019: Dong et al. (2004) showed that the equivalent width of the Fe line was 
1020: associated with column density, implying that the line could come from 
1021: the cold matter invoked in an absorbed model for active galactic nuclei (AGNs) NGC 2110 and NGC 7582.
1022: However, Asai et al. (2000) analyzed the Fe
1023: K$\alpha$ emission lines in the spectra of 20 LMXBs using
1024: {\sl ASCA} data, and found that the centroid energies of the Fe
1025: lines are almost independent of the source luminosity, and the
1026: detected equivalent width of the line also does not show a clear
1027: correlation with the source categories. 
1028: Although the physical parameters of determining the distinction of
1029: these categories has not been figured out,  the inclination angle,
1030: mass accretion rate, and the stellar magnetic field are the most
1031: relevant quantities. Thus, Asai et al. (2000) proposed that these 
1032: parameters may not play a key role in determining the Fe line strength.
1033: Unlike the conclusions of Asai et al. (2000), our results favor
1034: that the Fe emission line might be correlated with the accreting
1035: matter and come from the inner accretion disk.
1036: %
1037: 
1038: Nevertheless, we also noticed some discussions about the origin of
1039: the Fe emission line of other objects. For some black hole
1040: candidates, there is evidence that the observed Fe K$\alpha$ line
1041: originates in the innermost part of the accretion disk, close to the
1042: black hole (e.g., Fabian et al. 1989; Cui et al. 1998; Zycki et al. 1999).
1043: In addition, for AGNs, the higher
1044: resolution spectra available with the {\sl ASCA} appear to indicate that
1045: the line profiles are broad and asymmetrically skewed to lower
1046: energies, which is interpreted as evidence that the majority of the
1047: line emission origin from the inner accretion disk around the
1048: massive black hole (Nandra et al. 1997; Reynolds
1049: \& Begelman 1997). 
1050: Nandra et al. (2000) presented a spectral analysis of the Seyfert 1
1051: galaxy NGC 7469 with {\sl RXTE}, and found a significant correlation
1052: between the 2-10\,keV flux and the 6.4\,keV Fe K$\alpha$
1053: line, suggesting that the line emission comes from very nearby
1054: central regions.  
1055:  Our results imply that the Fe line emission region
1056: on the accretion disk of an X-ray binary may be located at the 
1057: same relative region as on the disk of an AGN.
1058: 
1059: %
1060: \section{Conclusion}
1061: \label{sect:conclusion}
1062: 
1063: In this paper, we analyzed the X-ray dip spectra of Cir X-1. The
1064: spectra can be well fitted with two blackbody components with the
1065: partial covering plus a third blackbody which is unaffected by
1066: partial covering.
1067: According to the spectral fitting results, the equivalent blackbody
1068: emission radii of the first (T$\sim$2\,keV), second (T$\sim$1\,keV) and third
1069: (T = 0.5$\sim$0.6\,keV) component are, respectively, 4-7\,km, 16-30\,km, 30-50\,km.
1070: The fluxes of the first and  second components are closely
1071: correlated, indicating that they represent emission from the inner regions.
1072: The emitting region of the
1073: coolest (3rd) component might come from the outer accretion disk.
1074: 
1075: %
1076: 
1077: A 6.4-6.7\,keV Fe emission line was detected in the  dip
1078: spectra.  
1079: %which is affected by the partial covering.
1080: %Furthermore, we find that, 
1081: During the long dips, the Fe line flux and fluxes
1082: of the first and  second components are correlated, and
1083: there is no significant correlation between the Fe line
1084: flux and the flux of the  third component, which suggests the Fe
1085: emission line comes from the region very close to the central
1086: object and is affected by the covering matter. During the short dip, 
1087: the Fe line flux is almost
1088: constant within the errors for all continuum spectra fluxes, which
1089: might be due to the fact that the line emission region lie in 
1090: the outer of the obscuring matter or the size of the obscuring matter
1091: is much smaller than that of the Fe line emitting region.
1092: 
1093: 
1094: \begin{acknowledgements}
1095: We would like to thank T.P. Li and S. N. Zhang for useful discussions.
1096: This research has made use of data obtained through the
1097: High Energy Astrophysics Science Archive Research Center
1098: Online Service, provided by the NASA/Goddard Space
1099: Flight Center. We acknowledge the {\sl RXTE} data teams at
1100: NASA/GSFC for their help. This work is subsidized by the
1101: Special Funds for Major State Basic Research Projects and
1102: by the National Natural Science Foundation of China.
1103: We are very grateful for the critic comments from the anonymous referee
1104:  that greatly improve the  quality of the paper.
1105: \end{acknowledgements}
1106: 
1107: \begin{thebibliography}{}
1108: \bibitem[]{}
1109: Asai K., Dotani T., Nagase F. et al. 2000, \apjs, 131, 571
1110: \bibitem[]{}
1111: Barnard R., Baluci{\' n}ska-Church M., Smale A.~P. et al. 2001, \aap, 380, 494
1112: 
1113: \bibitem[]{}
1114: Ba{\l}uci{\' n}ska-Church M., Church M.~J., Oosterbroek T. et al., 1999, \aap, 349, 495
1115: 
1116: \bibitem[]{}
1117: Belloni T., Mendez M., King A.~R. et al., 1997, \apj, 479, L145
1118: \bibitem[]{}
1119: Brandt W.~N., Fabian A.~C., Dotani T. et al., 1996,
1120: \mnras, 283, 1071
1121: 
1122: \bibitem[]{}
1123: Church M.~J.,  Balucinska-Church M., 1995, \aap, 300, 441
1124: 
1125: 
1126: \bibitem[]{}
1127: Church M.~J.,  Baluci{\' n}ska-Church M., 1996, Memorie della Societa
1128: Astronomica Italiana, 67, 395
1129: \bibitem[]{} Church M.~J., Dotani
1130: T., Balucinska-Church M. et al., 1997, \apj, 491, 388
1131: 
1132: 
1133: \bibitem[]{}
1134: Cui W., Ebisawa K., Dotani T. et al., 1998, \apj, 493, L75
1135: \bibitem[D{\'{\i}}az Trigo et al.(2006)]{2006A&A...445..179D} D{\'{\i}}az
1136: Trigo M., Parmar A.~N., Boirin L. et al., 2006, \aap, 445, 179
1137: \bibitem[]{}
1138: Ding G.Q., Qu J.L., Li T.P., 2006a, \aj, 131, 1693
1139: \bibitem[Ding et al.(2006)]{2006ApJ...645..576D} Ding G.~Q., Zhang S.~N.,
1140: Li T.~P. et al.,\ 2006b, \apj, 645, 576
1141: 
1142: \bibitem[]{}
1143: Done C., {\. Z}ycki P.~T., Smith D.~A., 2002, \mnras, 331, 453
1144: %\bibitem[]{}
1145: %Eggleton, P.~P.\ 1983, ApJ, 268, 368
1146: \bibitem[]{}
1147: Dong H., Xue S.J., Li C. et al., 2004, \chjaa, 4, 427
1148: \bibitem[]{}
1149: Fabian A.~C., Rees M.~J., Stella L. et al., 1989, \mnras, 238, 729
1150: \bibitem[]{}
1151: Goss W.~M., Mebold U., 1977, \mnras, 181, 255
1152: \bibitem[]{}
1153: Greiner J., Morgan E.~H., Remillard R.~A., 1996, \apj, 473,
1154: L107
1155: \bibitem[]{}
1156: Hirano T., Hayakawa S., Nagase F. et al., 1987, \pasj, 39, 619
1157: \bibitem[]{}
1158: Iaria R., Di Salvo T., Burderi L. et al., 2001, \apj, 561, 321
1159: \bibitem[Iaria et al.(2002)]{2002ApJ...567..503I} Iaria R., Di Salvo T.,
1160: Robba N.~R. et al., 2002, \apj, 567, 503
1161: \bibitem[]{}
1162: Inoue H., 1989, ESA SP-296: Two Topics in X-Ray Astronomy, Vol. 1: X Ray Binaries.~ Vol. 2: AGN and the
1163: X Ray Background, 109
1164: 
1165: \bibitem[]{}
1166: Mitsuda K., Inoue H., Nakamura N., Tanaka Y., 1989, \pasj, 41, 97
1167: 
1168: \bibitem[]{}
1169: Nandra K., Le T., George I.~M. et al., 2000, \apj, 544, 734
1170: \bibitem[]{}
1171: Nandra K., Mushotzky R.~F., Yaqoob T. et al., 1997, \mnras, 284, L7
1172: \bibitem[Oosterbroek et al.(2001)]{}
1173: Oosterbroek T., Parmar A.~N., Sidoli L. et al., 2001, \aap, 376, 532
1174: 
1175: \bibitem[]{}
1176: Reynolds C.~S., Begelman M.~C., 1997, \apj, 488, 109
1177: 
1178: \bibitem[]{}
1179: Shirey R.~E., Levine A.~M., Bradt H.~V., 1999, \apj, 524, 1048
1180: \bibitem[]{}
1181: Smale A.~P. et al., 1993, \apj, 410, 796
1182: \bibitem[]{}
1183: Stewart R.~T., Nelson G.~J., Penninx W. et al., 1991, \mnras, 253, 212
1184: 
1185: %\bibitem[]{}
1186: %Tauris T.~M., Fender R.~P., van den Heuvel E.~P.~J. et al.,\ 1999, \mnras,
1187: %310, 1165
1188: 
1189: \bibitem[]{}
1190: White N., Nagase F., Parmar A.~N., 1995, in X-ray Binaries, ed. W.~G.~H.
1191: Lewin J. van Paradijs,  E.~P.~J. van den Heuvel (Cambridge: Cambridge Univ. Press),1
1192: \bibitem[]{}
1193: White N.~E., Peacock A., Hasinger G. et al., 1986, \mnras, 218, 129
1194: 
1195: \bibitem[]{}
1196: White N.~E., Peacock A., Taylor B.~G., 1985, \apj, 296, 475
1197: \bibitem[]{}
1198: White N.~E., Stella L., Parmar A.~N., 1988, \apj, 324, 363
1199: 
1200: \bibitem[]{}
1201: Zycki P.~T., Done C., Smith D.~A., 1999, \mnras, 305, 231
1202: 
1203: \end{thebibliography}
1204: \label{lastpage}
1205: \end{document}
1206: %%==^..^============== the END of chjaa.tex ===================^_^==
1207: 
1208: 
1209: 
1210: 
1211: 
1212: 
1213: 
1214: 
1215: 
1216: 
1217: 
1218: 
1219: 
1220: 
1221: 
1222: