1:
2: % Norbert Schulz phone = (617) 258-5767
3:
4: \documentclass{aastex}
5: \usepackage{psfig}
6: % \usepackage{emulateapj5,apjfonts}
7:
8: \def\asca{{\it ASCA\/}}
9: \def\chandra{{\it Chandra\/}}
10: \def\einstein{{\it Einstein\/}}
11: \def\hst{{\it {\it HST}\/}}
12: \def\rosat{{\it ROSAT\/}}
13: \def\rxte{{\it RXTE\/}}
14: \def\sax{{\it BeppoSAX\/}}
15: \def\xmm{{\it XMM-Newton\/}}
16:
17: \def\cir{Cir~X-1}
18:
19: \def\todo{{\Huge $\bullet$}}
20:
21: \def\ltsima{$\; \buildrel < \over \sim \;$}
22: \def\simlt{\lower.5ex\hbox{\ltsima}}
23: \def\gtsima{$\; \buildrel > \over \sim \;$}
24: \def\simgt{\lower.5ex\hbox{\gtsima}}
25:
26: \slugcomment{Accepted for publication in The Astrophysical Journal Letters}
27:
28: \shorttitle{X-RAY P~CYGNI LINES FROM CIRCINUS~X-1}
29:
30: \shortauthors{BRANDT \& SCHULZ}
31:
32: \begin{document}
33:
34: \title{The Discovery of Broad P~Cygni X-ray Lines from Circinus~X-1\\
35: with the {\it CHANDRA\/} High Energy Transmission Grating Spectrometer}
36: \author{
37: W. N. Brandt\altaffilmark{1}
38: and
39: N. S. Schulz\altaffilmark{2}
40: }
41: \altaffiltext{1}{Department of Astronomy and Astrophysics, 525 Davey
42: Laboratory, The Pennsylvania State University, University Park, PA, 16802.}
43: \altaffiltext{2}{Center for Space Research, Massachusetts Institute of Technology,
44: Cambridge, MA 02139.}
45: \begin{abstract}
46: We present the first grating-resolution X-ray spectra of the X-ray binary
47: \cir, obtained with the High Energy Transmission Grating Spectrometer on
48: \chandra. These reveal a rich set of lines from
49: H-like and/or He-like Ne, Mg, Si, S and Fe detected with a high
50: signal-to-noise ratio. The lines are
51: broad ($\pm 2000$~km~s$^{-1}$) and show P~Cygni profiles. The absorption
52: components of the lines extend to low velocity, and they have about the
53: same widths and strengths as the corresponding emission components. The
54: widths of the X-ray P~Cygni lines are comparable to that of the broad
55: component of the strong, asymmetric H$\alpha$ line from \cir,
56: suggesting that the two phenomena may be related. We discuss
57: outflow models and propose that the P~Cygni profiles may
58: arise in the moderate temperature ($\sim 5\times 10^6$~K) region of
59: the wind from an X-ray heated accretion disk. This basic
60: picture strengthens the idea that the accretion disk in \cir\ is
61: viewed in a relatively edge-on manner, and it suggests that \cir\
62: is the X-ray binary analog of a Broad Absorption Line quasar.
63: \end{abstract}
64:
65: \keywords{
66: stars: individual (\cir) ---
67: stars: neutron ---
68: X-rays: stars ---
69: binaries: close ---
70: accretion: accretion disks ---
71: techniques: spectroscopic}
72:
73: % --------------------------------------------------------------------------------------
74:
75: \section{Introduction}
76: \label{intro}
77:
78: The most sensitive indicators of mass loss from cosmic sources are the
79: resonance absorption lines of abundant ions. In the ultraviolet, for example,
80: these lines have been used to study outflows in objects ranging from
81: stars to quasars. Prior to the launch of the
82: \chandra\ X-ray Observatory (e.g., Weisskopf et~al. 2000),
83: however, X-ray spectrometers generally lacked the resolution and
84: sensitivity needed to study resonance line absorption by the outflows
85: from most cosmic objects. Resonance absorption line studies
86: in the X-ray band offer a number of attractive features:
87: (1) the continuum driving the ionization is often directly visible,
88: (2) the atomic physics of the relevant ions is often relatively simple, and
89: (3) unlike ultraviolet photons, X-rays are not easily destroyed
90: in dusty environments.
91:
92: The mysterious X-ray binary \cir\ is thought to contain a neutron star
93: that can radiate at up to super-Eddington levels near the periastron
94: passage (`zero phase') of its 16.5-day orbital cycle
95: (e.g., Tennant, Fabian \& Shafer 1986; Inoue 1989), and its accretion
96: disk is probably viewed in a relatively edge-on manner (e.g., Brandt et~al. 1996).
97: The fact that it can radiate with high luminosity relative to its Eddington
98: luminosity ($L/L_{\rm Edd}$) makes it a natural system in which to expect
99: observable outflows, due to the larger amount of photon pressure available
100: per unit gravitational mass. Indeed, the
101: broad component (up to 2000~km~s$^{-1}$ FWHM) of the strong,
102: asymmetric H$\alpha$ line from \cir\ has been interpreted as arising in an
103: outflow from regions near the neutron star (e.g., Whelan et~al. 1977;
104: Johnston, Fender \& Wu 1999); the broad component of the H$\alpha$ line
105: is blueshifted relative to the narrow component.
106: Furthermore, the system appears to emit
107: radio jets and is embedded in a synchrotron nebula (e.g., Stewart et~al. 1993;
108: Fender et~al. 1998). Finally, we note a convergence of opinion that the
109: system is probably a low-mass X-ray binary based on
110: photometry and variability of the optical counterpart (e.g., Stewart et~al. 1991; Glass 1994),
111: its orbital period and apparent space velocity (Brandt \& Podsiadlowski 1996; Tauris et~al. 1999),
112: its correlated X-ray spectral and timing properties (e.g., Shirey, Bradt \& Levine 1999),
113: and the observed type~1 X-ray bursts (Tennant et~al. 1986).
114:
115: \nopagebreak[3]
116:
117: % This Letter describes the discovery of broad lines in the X-ray
118: % spectrum of \cir\ that show P~Cygni profiles. In \S2 we describe
119: % the observations and data analysis, and in \S3 we give a first
120: % discussion of the results. To our knowledge, the only previous
121: % observation of \cir\ with an X-ray grating spectrometer was with
122: % the Objective Grating Spectrometer (OGS) on \einstein\
123: % (Vrtilek et~al. 1991); \cir\ was in a low state during
124: % the OGS observation and was not detected.
125:
126: % --------------------------------------------------------------------------------------
127:
128: \section{Observations and Data Analysis}
129: \label{obser}
130:
131: \cir\ was observed using the High Energy Transmission Grating Spectrometer
132: (HETGS; C.R. Canizares et al., in preparation) on \chandra\ starting at
133: 22:09:50 UT on 2000 February 29 (see Figure~1).\footnote{For additional
134: information on the HETGS and ACIS see the \chandra\ Proposers' Observatory
135: Guide at http://asc.harvard.edu/udocs/docs.} The total integration time
136: was 32~ks, and the observation was continuous. The observation start time
137: corresponds to phase 0.99 using the ephemeris discussed by Glass (1994).
138:
139: The HETGS carries two types of transmission gratings, the
140: Medium Energy Grating (MEG) and the High Energy Grating (HEG). It allows
141: high-resolution spectroscopy from $\approx$~1.2--31~\AA\ ($\approx$~0.4--10~keV)
142: with a peak spectral resolution at
143: $\approx 12$~\AA\ of $\lambda/\Delta\lambda \approx 1000$ for HEG
144: 1st order. The resolution for higher orders improves by a factor of $n$ for the
145: $n$th order, but the spectral bandpass and efficiency are reduced for these
146: orders. In this analysis we mostly utilize the MEG and HEG 1st order as well
147: as the MEG 3rd order spectra.
148:
149: The dispersed spectra were recorded with an array of 6 Charge Coupled
150: Devices (CCDs) which are part of the Advanced CCD Imaging
151: Spectrometer (ACIS; G.P. Garmire et~al., in preparation).$^3$ To avoid CCD
152: frame `dropouts' due to telemetry saturation, we blocked the 0th order
153: image with a `spatial window' and shut off the two peripheral CCDs (S0 and S5).
154: The omission of the peripheral CCDs could be tolerated because \cir\ is heavily
155: absorbed by the interstellar column density
156: ($N_{\rm H}\approx 2\times 10^{22}$~cm$^{-2}$; e.g.,
157: Predehl \& Schmitt 1995) and offers hardly any flux
158: above $\approx 15$~\AA\ (e.g., Morrison \& McCammon 1983).
159: The brightness of \cir\ also required additional mitigation
160: efforts for `photon pile-up' effects. We applied a `subarray' during the
161: observation that reduced the CCD frame time to 1.7~s. The remaining pile-up
162: effects caused the 1st order spectra below $\approx 10$~\AA\ (MEG) and
163: $\approx 6$~\AA\ (HEG) to be depleted of photons, which then populate the
164: higher order spectra at shorter wavelengths. Below 7~\AA\ the MEG 1st order
165: spectrum is so depleted that we will not use it further in this analysis.
166: The MEG 3rd order spectrum is pile-up free from 2.3--9~\AA\ (1.4--5.4~keV).
167: Most of the lines we present below are not or are only moderately affected by
168: pile-up.
169:
170: Due to a misalignment of the applied subarray during the observation,
171: only half of the dispersed spectra were fully recorded; we only have
172: negative orders in the MEG and positive orders in the HEG complete.
173: The \chandra\ X-ray Center (CXC) provided aspect-corrected level~1 event
174: lists via its standard pipeline processing.
175:
176: % We reprocessed these data using the latest available data processing input
177: % products. We also removed events that resulted from either bad pixels or
178: % columns as well as from flaring background.
179:
180: The determination of the 0th order image position is crucial for the
181: calibration of the wavelength scales because it defines their zero
182: point. We calculated this position by fitting the dispersed images of the
183: MEG and HEG and by determining the intersections of these fits with the
184: 0th order `read-out trace' on the S3 CCD. The fits of the MEG and HEG were
185: consistent to within 0.3 detector pixels, which translates into a zero point
186: accuracy of 0.004~\AA\ for the MEG \,$-1$st order and 0.002~\AA\ for the
187: HEG \,$+1$st order. The overall wavelength calibration is good to $\approx 0.1$\%
188: and mostly depends on uncertainties in the positions of the CCD chip
189: gaps, which currently are $\pm 0.5$ pixels. For most of the spectral features
190: analyzed here there is only one gap involved leaving us with an overall
191: worst case uncertainty in the scale of
192: 0.008~\AA\ for MEG \,$-1$st order,
193: 0.005~\AA\ for HEG \,$+1$st order, and
194: 0.003~\AA\ for MEG \,$-3$rd order.
195:
196: We then processed the events into final event lists using CXC software, and
197: we used custom software and {\sc ftools} to produce our final grating spectra.
198: After standard grade selection, we have a total of
199: $1.35\times 10^6$ events in MEG \,$-1$st order,
200: $1.52\times 10^6$ events in HEG \,$+1$st order, and
201: $3.84\times 10^5$ events in MEG \,$-3$rd order.
202: We corrected our final grating spectra for instrumental effects
203: (e.g., energy dependent photon detection efficiency and gaps between the CCDs)
204: using aspect-corrected exposure maps.
205:
206: From the MEG \,$-3$rd order spectrum we determined a mean 2--8~keV flux of
207: $1.8\times 10^{-8}$~erg~cm$^{-2}$~s$^{-1}$. This was calculated by fitting
208: the \asca\ continuum model of Brandt et~al. (1996) to the 2.3--6.2~\AA\
209: (2--5.4~keV) spectrum (where pile-up is unimportant) and then extrapolating
210: the result to the 2--8~keV band. The observed flux is
211: within the range that has recently been seen by
212: \rxte\ (Shirey et~al. 1999) at zero phase. Even for a fairly small distance
213: of $\approx 6$~kpc (see \S3 of Stewart et~al. 1993 and
214: Case \& Bhattacharya 1998), the 2--8~keV luminosity is
215: $L_{2-8}\simgt 9.6\times 10^{37}$~erg~s$^{-1}$ (corrected for interstellar
216: absorption); hence for a $M=1.4$~$M_\odot$ neutron star with
217: $L_{\rm Edd}=1.3\times 10^{38} (M/M_\odot)$~erg~s$^{-1}$ we find
218: $L_{2-8}/L_{\rm Edd}\simgt 0.5$ (note there is certainly substantial
219: X-ray luminosity below 2~keV as well).
220: %
221: Figure~2 shows the resulting HEG \,$+1$st and MEG \,$-1$st order spectra. These
222: reveal a rich set of highly significant lines from H-like and/or He-like
223: Ne, Mg, Si, S and Fe. The lines are broad ($\pm 2000$~km~s$^{-1}$) and
224: show P~Cygni profiles (see Figure~3).
225:
226: Table~1 gives a list of the lines identified with P~Cygni profiles; many
227: of these have two or more independent detections (see Figure~3 for an
228: example). We used the line list presented in
229: Mewe, Gronenschild \& van den Oord (1985) to determine the line
230: wavelengths and transition types. Most measured wavelengths agree
231: to within 0.01~\AA\ with the predicted wavelengths. Some wavelengths
232: are measured to be slightly lower, which may be attributed to the fact
233: that some of the profiles are broadened and not entirely symmetric.
234: The `peak-to-valley' distances for the P~Cygni profiles are a
235: measure of the broadening of the lines. Table~1 shows that the
236: widths are quite significant (0.02--0.10~\AA), well above the
237: resolution of the HETGS.
238: %
239: % We have divided our observation into several segments and examined
240: % spectra for each of these.
241: %
242: While there is evidence for line variability over time, P~Cygni
243: profiles are persistently present.
244:
245: % visible in all spectra.
246:
247: % --------------------------------------------------------------------------------------
248:
249: \section{Discussion}
250:
251: % \subsection{Absorption Lines, Column Density and Ionization Parameter}
252: % \label{ionsec}
253:
254: The observed X-ray P~Cygni profiles are most naturally interpreted as arising
255: in a high-velocity outflow from the \cir\ system. Several properties of
256: the profiles deserve note.
257: %
258: First of all, the absorption troughs extend to low velocities; this
259: suggests that we are seeing the region in which the outflow is accelerated,
260: rather than, for example, just a shell of material that has already attained
261: its terminal velocity.
262: %
263: Secondly, the strengths of the emission and absorption components are
264: comparable when integrated over the whole observation.
265: %
266: Finally, the best-defined profiles look fairly smooth and continuous
267: down to the resolution limit of the HETGS; at least down to this limit
268: there is no evidence for strong shell-like density enhancements or
269: other inhomogeneities in the outflow.
270: %
271: The size of the \cir\ system is uncertain but is thought to be
272: $\sim 5\times 10^7$~km (e.g., Tauris et~al. 1999), so the characteristic
273: time for the outflow to cross the system is $\sim 0.5$~days.
274:
275: As mentioned in \S1, an outflow from the \cir\ system has already been
276: invoked to explain the broad component of its H$\alpha$
277: line (Johnston et~al. 1999). The widths of the X-ray P~Cygni profiles are
278: quite consistent with the FWHM of the broad component of H$\alpha$,
279: suggesting that these two phenomena may be related; ultraviolet P~Cygni
280: profiles and optical H$\alpha$ emission are often seen together from stars
281: with high mass loss rates (see \S2.3 of Lamers \& Cassinelli 1999).
282: Johnston et~al. (1999) argued that the broad H$\alpha$ line component
283: arises in a strong, anisotropic outflow driven by the radiation pressure
284: from the neutron star. In their model, this outflow must be optically
285: thick (to H$\alpha$ photons) to explain the broad blue wing of the
286: line (the redshifted emission being obscured by the outflow itself).
287:
288: The observed P~Cygni profiles are broadly consistent with such an outflow
289: model, although this outflow must be fairly dense to avoid becoming
290: completely ionized by the large X-ray luminosity of \cir.
291: In particular, the ionization parameter $\xi=L/nr^2$ must be
292: $\simlt 1000$~erg~cm~s$^{-1}$ so that the ions listed in
293: Table~1 can survive (see Kallman \& McCray 1982).\footnote{In
294: the definition of $\xi$, $L$ is the ionizing luminosity,
295: $n$ is the electron number density, and
296: $r$ is the distance to the ionizing radiation source.}
297: If we take the launching radius (i.e., onset radius)
298: of the outflow to be $r_{\rm launch}\approx 10^5$~km,
299: so that its observed terminal velocity, $v_{\rm t}$,
300: is comparable to the escape velocity, $v_{\rm esc}$,
301: at $r_{\rm launch}$, we require $n_{\rm launch}\simgt 10^{15}$~cm$^{-3}$
302: (we consider the plausible case of outflows launched from larger
303: radii where $v_{\rm t}>v_{\rm esc}$ below).
304: For comparison, the expected density of the accretion disk at this
305: radius is $\approx 2\times 10^{18}$~cm$^{-3}$, so the outflow-to-disk
306: density contrast seems plausible (compare with Figure~2 of
307: Raymond 1993).\footnote{We have calculated the
308: disk density following \S5.6 of Frank, King \& Raine (1992) with
309: a neutron star mass of $M=1.4$~$M_\odot$,
310: a mass accretion rate of $\dot M = 1.5\times 10^{18}$~gm~s$^{-1}$, and
311: a viscosity parameter of $\alpha=0.1$.}
312: To allow the survival of H~{\sc i} (for H$\alpha$)
313: in the outflow, $\xi\simlt 1$~erg~cm~s$^{-1}$ and hence
314: $n_{\rm H~I}\simgt 10^{18}$~cm$^{-3}$ is required; such a high density
315: outflow would be surprising, so it seems likely that most of the
316: H$\alpha$ emission is formed at a larger distance in the outflow
317: or in the accretion disk (see Mignani, Caraveo \& Bignami 1997).
318:
319: Based on the X-ray spectral variability properties of \cir,
320: Brandt et~al. (1996) argued that its accretion disk is flared and
321: viewed in a relatively edge-on manner. X-rays emitted in the inner
322: part of such a system can heat the surface of the disk farther
323: out, producing a corona and thermally driven wind
324: (e.g., Begelman, McKee \& Shields 1983). The outflow revealed
325: by our X-ray P~Cygni profiles might well be this wind, although
326: for this high $L/L_{\rm Edd}$ system radiation pressure acting
327: on lines (in addition to Compton heating) is likely to be important.
328: %
329: Raymond (1993) has examined
330: atomic processes in the context of this scenario, and his
331: calculations show that at distances of $\approx 10^5$--$10^6$~km
332: there is a significant (subtending $\sim 3$--20$^\circ$), moderate
333: temperature ($\sim 5\times 10^6$~K) region where atomic heating and
334: cooling processes dominate over Compton processes. Significant line
335: emission is expected from this region, and in fact most of the lines
336: we list in Table~1 are those predicted to be strong by Raymond (1993).
337: Thus, our favored interpretation for the observed X-ray P~Cygni
338: profiles is that they arise in the intermediate temperature region
339: of the wind from an accretion disk viewed in a relatively edge-on
340: manner. \cir\ then becomes an X-ray binary analog of a Broad Absorption
341: Line quasar.
342: %
343: At smaller radii, wind and coronal material
344: are likely to be heated to the Compton temperature and thus
345: completely ionized. An appealing physical possibility is that the
346: electron scatterer discussed in \S4.1 of Brandt et~al. (1996) is
347: just this material. Similar highly ionized gas has been
348: invoked in some models for Broad Absorption Line quasars
349: (e.g., the `hitchhiking gas' of Murray et~al. 1995).
350:
351: Despite the general attractiveness of the above picture, we
352: must note a potential difficulty: for a launching radius with
353: $v_{\rm t}\approx v_{\rm esc}$, the quantity
354: $N_{\rm H,launch}=\int_{r_{\rm launch}}^{\infty}n\,dr
355: =n_{\rm launch} r_{\rm launch}\simgt 10^{25}$~cm$^{-2}$
356: is so large that the implied column density through the wind
357: is optically thick to electron scattering [here we assume
358: that the wind has a radial extent $\gg r_{\rm launch}$ and
359: that $n=n_{\rm launch} (r_{\rm launch}/r)^2$]. Most line
360: photons attempting to traverse the wind would then be Compton
361: scattered out of the line.
362: %
363: Shielding of the wind from the full X-ray continuum may
364: alleviate this problem by allowing a significant
365: reduction in $n_{\rm launch}$ (see Begelman \& McKee 1983).
366: %
367: Alternatively, increasing $r_{\rm launch}$ helps because
368: $N_{\rm H,launch}\propto r_{\rm launch}^{-1}$ when we
369: require $\xi\simlt 1000$~erg~cm~s$^{-1}$ and thus
370: $n_{\rm launch}\simgt L/(r^2_{\rm launch}~1000~{\rm erg~cm~s^{-1}})$;
371: in this case significant radiation pressure driving may be needed
372: since $v_{\rm t}>v_{\rm esc}$. $r_{\rm launch}$ cannot be
373: $\gg 10^{6}$~km as the disk probably has an outer radius
374: $\simlt 3\times 10^{6}$~km (see Figure~3 of Tauris et~al. 1999).
375: %
376: Finally, clumping of the wind might also help because when
377: clumping is present $\xi=Lf/nr^2$ where $f$ is the volume
378: filling factor.
379: %
380: We note that Iaria et~al. (2000) have recently found evidence
381: for a large ionized column density ($\sim 10^{24}$~cm$^{-2}$)
382: along the line of sight even when \cir\ is radiating at
383: high luminosity; this material may be the same that makes
384: the P~Cygni lines.
385: %
386: We are presently examining these issues in further detail.
387:
388: While we cannot rigorously rule out the possibility that the
389: X-ray P~Cygni lines arise in a wind from the companion star,
390: we consider this unlikely. First of all, the large wind
391: velocity ($\pm 2000$~km~s$^{-1}$) implied by the profiles
392: would require a high-mass secondary for the system
393: (for an overview see \S2.7 of Lamers \& Cassinelli 1999),
394: provided the presence of the radiating compact object does not
395: lead to a substantially faster wind from the companion
396: than would otherwise be expected.
397: However, as mentioned in \S1, the bulk of the evidence suggests a
398: low-mass X-ray binary nature. Furthermore, the large X-ray luminosity of
399: \cir\ should completely ionize an O~star wind out to
400: $\simgt 5\times 10^6$~km (compare with Boroson et~al. 1999),
401: while the observed P~Cygni profiles suggest that we are seeing
402: the acceleration region of the outflow.
403:
404: % A future paper will present detailed spectral and
405: % timing analyses of the X-ray P~Cygni profiles,
406: % a complete set of the spectral features observed
407: % (e.g., weaker fluorescent emission lines), and
408: % fitting of the spectral continuum.
409: % %
410: % We also intend to make additional HETGS observations to study how
411: % the P~Cygni profiles change with orbital phase and luminosity;
412: % this should further clarify their origin.
413:
414: To our knowledge, these are the first reported X-ray P~Cygni profiles
415: from an X-ray binary. Hopefully X-ray P~Cygni profiles
416: will be identified and studied in other systems to provide
417: geometrical and physical insight into the flows of material near
418: Galactic compact objects.
419:
420: % --------------------------------------------------------------------------------------
421:
422: \acknowledgments
423: We thank all the members of the \chandra\ team for their enormous efforts,
424: and we thank
425: J. Chiang,
426: A.C. Fabian,
427: S.C. Gallagher,
428: S. Kaspi,
429: R.A. Wade, and
430: an anonymous referee
431: for helpful discussions.
432: %
433: We gratefully acknowledge the financial support of
434: CXC grant GO0-1041X (WNB, NSS),
435: the Alfred P. Sloan Foundation (WNB), and
436: Smithsonian Astrophysical Observatory contract SV1-61010 for the CXC (NSS).
437:
438: % --------------------------------------------------------------------------------------
439:
440: \begin{thebibliography}{}
441:
442: \bibitem[]{}
443: Begelman, M.C., McKee, C.F. \& Shields, G.A. 1983, ApJ, 271, 70
444:
445: \bibitem[]{}
446: Begelman, M.C. \& McKee, C.F. 1983, ApJ, 271, 89
447:
448: \bibitem[]{}
449: Boroson, B., Kallman, T., McCray, R., Vrtilek, S. \& Raymond, J.
450: 1999, ApJ, 519, 191
451:
452: \bibitem[]{}
453: Brandt, W.N. \& Podsiadlowski, Ph. 1995, MNRAS, 274, 461
454:
455: \bibitem[]{}
456: Brandt, W.N., Fabian, A.C., Dotani, T., Nagase, F., Inoue, H.,
457: Kotani, T. \& Segawa, Y. 1996, MNRAS, 283, 1071
458:
459: \bibitem[]{}
460: Case, G.L. \& Bhattacharya, D. 1998, ApJ, 504, 761
461:
462: \bibitem[]{}
463: Fender, R., Spencer, R., Tzioumis, T., Wu, K., van der Klis, M.,
464: van Paradijs, J. \& Johnston, H. 1998, ApJ, 506, L121
465:
466: \bibitem[]{}
467: Frank, J., King, A. \& Raine, D.
468: 1992,
469: Accretion Power in Astrophysics
470: (Cambridge Univ. Press, Cambridge)
471:
472: \bibitem[]{}
473: Glass, I. 1994, MNRAS, 268, 742
474:
475: \bibitem[]{}
476: Iaria, R., Burderi, L., Di~Salvo, T., La~Barbera, A. \& Robba, N. R.
477: 2000, ApJ, in press (astro-ph/0009183)
478:
479: \bibitem[]{}
480: Inoue, H. 1989,
481: in Proc. 23rd ESLAB Symp. on Two Topics in X-ray Astronomy,
482: ed. N.E. White, J. Hunt \& B. Battrick
483: (Paris: ESA Publications), 109
484:
485: \bibitem[]{}
486: Johnston, H.M., Fender, R. \& Wu, K. 1999, MNRAS, 308, 415
487:
488: \bibitem[]{}
489: Kallman, T.R. \& McCray, R. 1982, ApJS, 50, 263
490:
491: \bibitem[]{}
492: Lamers, H.J.G.L.M. \& Cassinelli, J.P.
493: 1999,
494: Introduction to Stellar Winds
495: (Cambridge Univ. Press, Cambridge)
496:
497: \bibitem[]{}
498: Mewe, R., Gronenschild, E.H.B.M. \& van den Oord, G.H.J.
499: 1985, A\&AS, 62, 197
500:
501: \bibitem[]{}
502: Mignani, R., Caraveo, P.A. \& Bignami, G.F. 1997, A\&A, 323, 797
503:
504: \bibitem[]{}
505: Morrison, R. \& McCammon, D.
506: 1983, ApJ, 270, 119
507:
508: \bibitem[]{}
509: Murray, N., Chiang, J., Grossman, S.A. \& Voit, G.M.
510: 1995, ApJ, 451, 498
511:
512: \bibitem[]{}
513: Predehl, P. \& Schmitt, J.H.M.M. 1995, A\&A, 293, 889
514:
515: \bibitem[]{}
516: Raymond, J.C. 1993, ApJ, 412, 267
517:
518: \bibitem[]{}
519: Shirey, R.E., Bradt, H.V. \& Levine, A.M. 1999, ApJ, 517, 472
520:
521: \bibitem[]{}
522: Stewart, R.T., Nelson, G.J., Penninx, W., Kitamoto, S., Miyamoto, S. \& Nicholson, G.D.
523: 1991, MNRAS, 253, 212
524:
525: \bibitem[]{}
526: Stewart, R.T., Caswell, J.L., Haynes, R.F. \& Nelson, G.J., 1993, MNRAS, 261, 593
527:
528: \bibitem[]{}
529: Tauris, T.M., Fender, R.P., van den Heuvel, E.P.J., Johnston, H.M. \& Wu, K.
530: 1999, MNRAS, 310, 1165
531:
532: \bibitem[]{}
533: Tennant, A.F., Fabian, A.C. \& Shafer, R.A. 1986, MNRAS, 221, 27P
534:
535: % \bibitem[]{}
536: % Vrtilek, S.D., McClintock, J.E., Seward, F.D., Kahn, S.M. \& Wargelin, B.J.
537: % 1991, ApJS, 76, 1127
538:
539: \bibitem[]{}
540: Weisskopf, M.C., Tananbaum, H.D., Van Speybroeck, L.P. \& O'Dell, S.L.
541: 2000, Proc. SPIE, in press (astro-ph/0004127)
542:
543: \bibitem[]{}
544: Whelan, J.A.J., et~al. 1977, MNRAS, 181, 259
545:
546: % \bibitem[]{}
547: % Wickramasinghe, D.T., Bicknell, G.V. \& Ferrario, L.
548: % 1997,
549: % Accretion Phenomena and Related Outflows
550: % (ASP Press, San Francisco)
551:
552: \end{thebibliography}
553:
554: \newpage
555:
556: %\begin{deluxetable}{lccc}
557: %\tablecolumns{4}
558: %\tabletypesize{\small}
559: %\tabletypesize{\footnotesize}
560: %\tabletypesize{\scriptsize}
561: %\tablewidth{0pt}
562: %\tablecaption{P~Cygni lines from \cir
563: %\label{colden} }
564: %\tablehead{
565:
566: \vbox{
567: \footnotesize
568: \begin{center}
569: {\sc TABLE 1\\
570: X-ray P~Cygni lines from \cir}
571: \vskip 4pt
572: \begin{tabular}{llccc}
573: \hline
574: \hline
575: %
576: {Ion} &
577: {Isoelectronic seq.} &
578: {Predicted $\lambda$} &
579: {Measured $\lambda$} &
580: {$\Delta\lambda$} \\
581: %
582: {Name} &
583: {and transition$^{\rm a}$} &
584: {(\AA)$^{\rm b}$} &
585: {(\AA)$^{\rm c}$} &
586: {(\AA)$^{\rm d}$} \\
587: %
588: \hline
589: \ion{Fe}{26} & H-like (Ly$\alpha$) & 1.78 & 1.78 & 0.02 \\
590: \ion{Fe}{25} & He-like & 1.85 & 1.85 & 0.02 \\
591: \ion{S}{16} & H-like (Ly$\alpha$) & 4.73 & 4.72 & 0.03 \\
592: \ion{S}{15} & He-like & 5.04 & 5.04 & 0.03 \\
593: \ion{Si}{14} & H-like (Ly$\beta$) & 5.22 & 5.21 & 0.03 \\
594: \ion{Si}{14} & H-like (Ly$\alpha$) & 6.18 & 6.17 & 0.03 \\
595: \ion{Mg}{12} & H-like (Ly$\gamma$) & 6.74 & 6.73 & 0.04 \\
596: \ion{Mg}{12} & H-like (Ly$\alpha$) & 8.42 & 8.40 & 0.06 \\
597: \ion{Ne}{10} & H-like (Ly$\delta$) & 9.48 & 9.48 & 0.09 \\
598: \ion{Fe}{24} & Li-like & 10.63 &10.63 & 0.10 \\
599: \ion{Ne}{10} & H-like (Ly$\alpha$) & 12.13 &12.11 & 0.06 \\
600: %
601: % \ion{Mg}{11} & 7.85 & 7.86 & 0.04 \\
602: %
603: \hline
604: \vspace*{0.02in}
605: \end{tabular}
606: \parbox{3.2in}{
607: % \parbox{2.5in}{
608: \small\baselineskip 9pt
609: \footnotesize
610: \indent
611: $\rm ^a${We list the transition here if it can be written compactly; for the other
612: transitions see Tables~1 and 3 of Mewe et~al. (1985).} \\
613: $\rm ^b${From Mewe et~al. (1985).} \\
614: $\rm ^c${Measured intersection of the P~Cygni profile (between the emission maximum
615: and absorption minimum) with a local continuum fit. The measurement accuracy
616: corresponds to about 1 data bin (i.e., $\pm 0.0025$~\AA).} \\
617: $\rm ^d${Measured `peak-to-valley' distances for the P~Cygni profiles.}
618: }
619: \end{center}
620: \setcounter{table}{1}
621: \normalsize
622: \centerline{}
623: }
624:
625: \newpage
626:
627:
628:
629: \begin{figure}[t!]
630: \hbox{
631: \hspace*{0.6 in} \psfig{figure=fig1.ps,height=5.0truein,width=5.0truein,angle=0.}
632: }
633: \caption{
634: \rxte\ All-Sky Monitor (ASM) light curve of \cir\
635: around the time of our \chandra\ observation. The
636: start and stop times of our \chandra\ observation are denoted
637: by dotted vertical lines. The downward pointing arrow shows the
638: zero-phase time of \cir. For comparison, the count rate of the
639: Crab in the \rxte\ ASM is $\approx 75$~count~s$^{-1}$.
640: }
641: \end{figure}
642:
643: \begin{figure}[t!]
644: \hbox{
645: \hspace*{0.6 in} \psfig{figure=fig2a.ps,height=5.0truein,width=5.0truein,angle=0.}
646: }
647: \caption{
648: Exposure corrected (a) HEG \,$+1$st order and (b) MEG \,$-1$st order
649: spectra of \cir\ with a spectral binning of 0.005~\AA. These spectra have been
650: smoothed with a `boxcar' filter of width 3 bins, and they have not been corrected
651: for the effects of interstellar absorption. We have marked
652: all the P~Cygni type spectral lines that are detected with
653: $>5\sigma$ significance and that are mostly free from contamination effects
654: (we do not detect a P~Cygni profile from Fe~{\sc i} so this line is starred).
655: The feature at 4.31~\AA\ corresponds to S~XV, but it is not marked because
656: its amplitude is strongly affected by uncertainties in the CCD gap correction.
657: }
658: \end{figure}
659:
660:
661:
662: \begin{figure}[t!]
663: \hbox{
664: \hspace*{0.6 in} \psfig{figure=fig2b.ps,height=5.0truein,width=5.0truein,angle=0.}
665: }
666: % \caption{}
667: \end{figure}
668:
669:
670: \begin{figure}[t!]
671: \hbox{
672: \hspace*{1 in} \psfig{figure=fig3.ps,height=5.0truein,width=4.0truein,angle=0.}
673: }
674: \caption{
675: Velocity spectra showing the details of a few of the strongest
676: X-ray P~Cygni profiles seen from \cir. We show the independent measurements
677: of the Si~{\sc xiv} line from both the HEG \,$+1$st order and MEG \,$-3$rd order
678: spectra. Typical bins in this figure have 200--1200 counts, and these spectra
679: have not been smoothed.
680: %
681: We list the relevant velocity resolution in each panel. The lines are
682: clearly broader than the instrumental resolution with velocities of
683: $\pm 2000$~km~s$^{-1}$, although the Fe~{\sc xxv} line is broadened by
684: the instrument.
685: %
686: We have taken zero velocity to correspond to the laboratory rest wavelength,
687: since the radial velocity of \cir\ is not well established
688: (Johnston et~al. 1999 and H.M. Johnston 2000, private communication).
689: }
690: \end{figure}
691:
692: % \vspace*{0.2in}
693:
694: % --------------------------------------------------------------------------------------
695:
696:
697: \end{document}
698: