1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{psfig}
3:
4: \newcommand\mdot {\hbox {${\dot M}$}}
5: \newcommand\sz {$S_{\rm z}$}
6: \newcommand\mzon {M$_{\odot}$}
7: \newcommand\pp {$\pm$}
8: \newcommand\pers {s$^{-1}$}
9: \newcommand\micros {$\mu$s}
10:
11: \def\degr{\hbox{$^\circ$}}
12: \newcommand\Lunit {erg s$^{-1}$}
13: \newcommand\funit {erg cm$^{-2}$ s$^{-1}$}
14: \def\degr{\hbox{$^\circ$}}
15:
16: \def\la{\mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
17: $\displaystyle##$\hfil\cr<\cr\noalign{\vskip1.5pt}\sim\cr}}}
18: {\vcenter{\offinterlineskip\halign{\hfil$\textstyle##$\hfil\cr<\cr
19: \noalign{\vskip1.0pt}\sim\cr}}}
20: {\vcenter{\offinterlineskip\halign{\hfil$\scriptstyle##$\hfil\cr<\cr
21: \noalign{\vskip0.5pt}\sim\cr}}}
22: {\vcenter{\offinterlineskip\halign{\hfil$\scriptscriptstyle##$\hfil
23: \cr<\cr\noalign{\vskip0.5pt}\sim\cr}}}}}
24:
25:
26: \righthead{GRO J1744--28 in quiescence}
27: \slugcomment{Re-submitted to ApJ Letters: February 12, 2002}
28:
29: \begin{document}
30:
31: \title{A {\itshape Chandra} observation of GRO J1744--28: the bursting
32: pulsar in quiescence}
33:
34: \author{Rudy Wijnands\altaffilmark{1,2}, Q. Daniel Wang\altaffilmark{3}}
35:
36: \altaffiltext{1}{Center for Space Research, Massachusetts Institute of
37: Technology, 77 Massachusetts Avenue, Cambridge, MA 02139-4307, USA;
38: rudy@space.mit.edu}
39:
40: \altaffiltext{2}{Chandra Fellow}
41:
42:
43: \altaffiltext{3}{Astronomy Department, University of Massachusetts,
44: Amherst, MA 01003, USA}
45:
46: \begin{abstract}
47: We present a {\it Chandra}/ACIS-I observation of GRO J1744--28. We
48: detected a source at a position of R.A = 17$^h$ 44$^m$ 33.09$^s$ and
49: Dec. = --28\degr~44$'$ 27.0$''$ (J2000.0; with a 1$\sigma$ error of
50: $\sim$0.8 arcseconds), consistent with both {\it ROSAT} and
51: interplanetary network localizations of GRO J1744--28 when it was in
52: outburst. This makes it likely that we have detected the quiescent
53: X-ray counterpart of GRO J1744--28. Our {\it Chandra} position
54: demonstrates that the previously proposed infrared counterpart is not
55: related to GRO J1744--28. The 0.5--10 keV luminosity of the source is
56: $2 - 4 \times10^{33}$ \Lunit~(assuming the source is near the Galactic
57: center at a distance of 8 kpc). We discuss our results in the context
58: of the quiescent X-ray emission of pulsating and non-pulsating neutron
59: star X-ray transients.
60:
61: \end{abstract}
62:
63:
64: \keywords{pulsars: individual (GRO J1744--28) --- stars: neutron ---
65: X-rays: stars}
66:
67: \section{Introduction \label{section:intro}}
68:
69: X-ray transients sporadically exhibit very bright outbursts during
70: which their X-ray luminosity can be as high as $10^{36-39}$
71: \Lunit. However, most of their time they are in their quiescent state
72: during which they emit X-rays only at a level of $10^{30-34}$ \Lunit.
73: The mechanisms behind this quiescent X-ray emission are still not
74: understood (see, e.g., Menou et al. 1999; Campana \& Stella 2000;
75: Bildsten \& Rutledge 2002). The most dominant model for the quiescent
76: properties of non-pulsating neutron star transients, which contain a
77: neutron star with a very low magnetic field ($<10^{10}$ Gauss), is the
78: one which assumes that the X-rays below a few keV are due to the
79: cooling of the neutron star after the accretion has stopped (see,
80: e.g., Brown, Bildsten, \& Rutledge 1998 and references therein). In
81: the quiescent X-ray spectra of several of those systems an extra power
82: law component above a few keV is present (see, e.g., Asai et al. 1996,
83: 1998; Campana et al. 1998a; Rutledge et al. 2001), but the nature of
84: this component is even more unclear (see, e.g., Campana \& Stella 2000
85: for a discussion).
86:
87: Detailed studies of the quiescent emission from transient X-ray
88: pulsars (with a neutron star magnetic field strength of $>10^{11}$
89: Gauss) have been inhibited by the lack of detected systems. So far,
90: only two X-ray pulsars have been detected in quiescence, A 0535+26
91: (Negueruela et al. 2000) and 4U 0155+63 (Campana et al. 2001). Their
92: X-ray luminosities were consistent with the predictions of the cooling
93: neutron star model for the non-pulsating systems, suggesting that also
94: this model might apply to the pulsating ones. However, due to the low
95: statistics of the data a detailed testing of the model could not be
96: performed. To get more insight in the quiescent emission of X-ray
97: pulsars those systems have to be observed with higher sensitivity
98: instruments like {\it Chandra} or {\it XMM-Newton}. Furthermore, more
99: quiescent X-ray pulsars have to be detected to determine if all of
100: those systems are consistent with the cooling model or that some
101: systems might have different properties which would suggest
102: alternative X-ray production mechanisms (e.g., accretion down to the
103: magnetospheric radius; Stella et al. 1994; Corbet 1996). A good
104: candidate for such studies is the ``bursting pulsar'' GRO J1744--28.
105:
106: GRO J1744--28 was discovered in December 1995 with the Burst and
107: Transient Source Experiment (BATSE) aboard the {\it Compton Gamma-Ray
108: Observatory} ({\it CGRO}; Fishman et al. 1995; Kouveliotou et
109: al. 1996). The source exhibited rapidly repeating, very bright X-ray
110: bursts which are likely due to accretion disk instabilities (e.g.,
111: Lewin et al. 1996). The source also exhibited pulsed emission with a
112: pulsation frequency of 2.1 Hz (Finger et al. 1996). Its bursting and
113: pulsating nature lead to the source being called the bursting
114: pulsar. So far, two major outbursts have been detected, one starting
115: in December 1995 and the other one a year later in December 1996. The
116: latter one ended in April 1997 (see, e.g., Woods et al. 1999).
117:
118: Augusteijn et al. (1997) likely detected GRO J1744--28 using a {\it
119: ROSAT} observation performed in March 1996. This observation showed a
120: bright source with a luminosity of $\sim 2 \times 10^{37}$ \Lunit~(for
121: an assumed distance\footnote{The distance to the source is unknown but
122: the high column density measured by Nishiuchi et al. (1999) and the
123: proximity of the source on the sky to the center of the Galaxy make a
124: large distance likely.} of 8 kpc; 0.1--2.4 keV). Archival {\it ROSAT}
125: observations did not detect a source on this position with an upper
126: limit on the luminosity of a few times $10^{33}$ \Lunit~(0.1--2.4 keV;
127: Augusteijn et al. 1997). This transient nature of the source makes it
128: very likely that indeed the {\it ROSAT} source is GRO J1744--28
129: despite that no bursts or pulsations were detected (note that the
130: large {\it ROSAT} upper limit on the pulsations obtained by Augusteijn
131: et al. 1997 was completely consistent with the strength of the
132: pulsations as measured with BATSE and the {\it Rossi X-ray Timing
133: Explorer} [{\it RXTE}]). Localization of GRO J1744--28 by
134: triangulating the data obtained for this source with {\it Ulysses} and
135: BATSE (part of the interplanetary network [IPN]) resulted in a
136: position (Hurley et al. 2000) which partly overlapped that of the {\it
137: ROSAT} error circle, confirming the identification of the {\it ROSAT}
138: source with GRO J1744--28. Cole et al. (1997) and Augusteijn et
139: al. (1997) identified a possible infrared counterpart at the edge of
140: the {\it ROSAT} error circle, although its enigmatic properties
141: spurred the suggestion that it might be an instrumental artifact
142: (Augusteijn et al. 1997; but see Cole et al. 1997).
143:
144: Here we present a {\it Chandra}/ACIS-I observation of the region
145: containing GRO J1744--28. We discovered a weak source near the center
146: of the {\it ROSAT} error circle, which is likely the quiescent
147: counterpart of the {\it ROSAT} transient and therefore likely of GRO
148: J1744--28.
149:
150:
151: \section{Observation, analysis, and results}
152:
153: GRO J1744--28 was in the field of view ($\sim$7.1 arcminutes off-axis)
154: during one of the observations which were obtained as part of the {\it
155: Chandra} survey of the Galactic Center region (Wang, Gotthelf, \& Lang
156: 2002). This particular observation was performed on 2001 July 18 17:37
157: - 20:49 UT with an exposure time of $\sim10.6$ ksec. The ACIS-I
158: instrument was used during this observation. To limit the telemetry
159: rate only those photons with energy above 1 keV were transmitted to
160: Earth. The data were analysed using the analysing packet CIAO, version
161: 2.2.1, and the threats listed on the CIAO web pages\footnote{Available
162: at http://asc.harvard.edu/ciao/}.
163:
164: The resulting image of the region near GRO J1744--28 is shown in
165: Figure~\ref{fig:image}. We detected one source in the {\it ROSAT}
166: error circle (Augusteijn et al. 1997) and the IPN error ellipse
167: (Hurley et al. 2000) of GRO J1744--28. This strongly suggests that
168: indeed we have detected the quiescent counterpart of GRO J1744--28.
169: The best source position (as obtained with the CIAO tool WAVDETECT) is
170: R.A. = 17$^h$ 44$^m$ 33.09$^s$ and Dec. = --28\degr~44$'$ 27.0$''$
171: (for J2000.0). Due to the low number of detected photons the
172: statistical error on the source position is 0.3 arcseconds. However,
173: the satellite pointing error is approximately 0.7 arcseconds
174: (1$\sigma$; Aldcroft et al. 2000) and dominates the positional
175: inaccuracy. The proposed infrared counterpart (Cole et al. 1997;
176: Augusteijn et al. 1997) is not consistent with our {\it Chandra}
177: position.
178:
179:
180: The elongated structure of the detected source in
181: Figure~\ref{fig:image} is due to the extended point-spread-function
182: for a point source approximately 7 arcminutes off-axis. We detected
183: 52\pp8 counts (corrected for background) from the source position,
184: resulting in a count rate of 4.9\pp0.7 $\times10^{-3}$ counts
185: s$^{-1}$. The source spectrum was extracted using a circle with a
186: radius of 10$''$ on the source position\footnote{For a $\sim7'$ off
187: axis source an extraction radius of 10$''$ ensures that virtually all
188: source photons are extracted and that no contamination of source
189: photons occurs in the background region.}. The background data were
190: obtained by using an annulus on the same position with an inner radius
191: of 10$''$ and an outer one of 30$''$. The data were rebinned using the
192: FTOOLS routine grppha into bins with a minimum of 5 counts per bin. We
193: employed both the $\chi^2$ and CASH (Cash 1979) statistics to fit the
194: data. Both methods give very similar results and we will only discuss
195: the results obtained with the $\chi^2$ method. The obtained spectrum
196: was fitted using XSPEC version 11.1.0 (Arnaud 1996).
197:
198: The spectrum (Fig.~\ref{fig:spectra}) was of poor quality and every
199: single-component model provided an acceptable fit ($\chi^2/dof \la
200: 1$). The column density could not be constrained and was fixed to the
201: value as determined using {\it ASCA} data ($N_{\rm H} \sim 5.5 \times
202: 10^{22}$ cm$^{-2}$; Nishiuchi et al. 1999) during times when the
203: source was in outburst. A pure power-law model resulted in a photon
204: index of 2.4\pp0.7 (all fit parameters are for a 95\% confidence
205: level) and a flux of $5.3\times10^{-13}$ \funit~(0.5--10 keV;
206: unabsorbed; all subsequent fluxes are for this energy range and
207: unabsorbed). A blackbody model resulted in a temperature $kT$ of
208: 0.8\pp0.2 keV, with a radius of the emitting region of only
209: 0.2$^{+1.6}_{-0.9}$ km (assuming a distance of 8 kpc), and a flux of
210: $2.5\times10^{-13}$ \funit. When fitting a neutron star hydrogen
211: atmosphere model (Zavlin, Pavlov, \& Shibanov 1996) to the data, the
212: radius of the emitting region could not be constrained and was fixed
213: to 10 km. The obtained temperature at infinity $kT_\infty$ was
214: 0.4\pp0.2 keV and the resulting flux is $3.0\times10^{-13}$
215: \funit. The uncertainties in which spectral model to use resulted in a
216: possible range for the flux of $2.5 - 5.3 \times 10^{-13}$ \funit. If
217: the source is indeed near the Galactic center at a distance of $\sim8$
218: kpc, then the 0.5--10 keV luminosity would be in the range of $2 - 4
219: \times 10^{33}$ \Lunit.
220:
221:
222: When the quiescent spectrum of non-pulsating neutron star X-ray
223: transients are fit with a single component model consisting of either
224: a blackbody or a neutron star atmosphere model, the obtained
225: temperatures are usually 0.2--0.3 keV (using a blackbody model; e.g.,
226: Bildsten \& Rutledge 2002 and references therein) or $\sim0.1$ keV
227: (using a neutron star atmosphere model). Our spectrum of GRO J1744--28
228: is not consistent with such low temperatures ($\chi^2/dof > 3$): both
229: the blackbody and the neutron star atmosphere model fall-off more
230: rapidly at energies above 1 keV than our data. However, several of
231: those non-pulsating systems have displayed a composite quiescent X-ray
232: spectrum, with a soft, most likely thermal component below $\sim1$ keV
233: and a hard, power-law type component at higher energies (e.g., Asai et
234: al. 1996, 1998). We have fitted our data using such a composite
235: spectrum consisting of either a blackbody or a neutron star atmosphere
236: model for the soft component and a power-law with photon index of 1 or
237: 2 for the hard component. However, due to the low number of photons
238: detected and the high column density towards the source, no
239: constraints could be set on the temperature of the soft component. The
240: data are fully consistent with the temperatures observed for other
241: systems (e.g., a blackbody temperature of 0.2--0.3 keV) and with an
242: upper limit on the thermal flux of $3 \times 10^{-13}$ \funit~(0.5--10
243: keV; for a $kT$ of 0.3 keV) or $9 \times 10^{-13}$ \funit~(for a $kT$
244: of 0.2 keV).
245:
246: The detected source was the only one present in the {\it ROSAT} error
247: circle and the IPN ellipse. Hands et al. (2002) reported on {\it
248: XMM-Newton} observations performed on the Galactic plane to study the
249: Galactic X-ray point source population. Using their log $N$ -- log $S$
250: curve and a 2--6 keV flux of $\sim 1 - 2\times10^{-13}$ \funit~for GRO
251: J1744--28, we estimate that about 10 to 20 sources per square degree
252: can be detected with the same flux. This gives a probability of $\sim1
253: - 2\times10^{-4}$ that a random field source would fall in the surface
254: area traced out by the intersection of the {\it ROSAT} error circle
255: and the IPN error ellipse. The source density will likely be higher in
256: the Galactic center region which would increase this probability. But
257: the calculated probability can be used as a first approximation and it
258: indicates that the detected source is likely GRO J1744--28. However,
259: if the detected {\it Chandra} source is not GRO J1744--28, then the
260: flux upper limit on this source would be about $2 - 5 \times 10^{-14}$
261: \funit~(depending on which spectral model is assumed). Our
262: observation had a time resolution of only $\sim3.2$ seconds, which did
263: not allow us to search for the 2.1 Hz pulsations. The low number of
264: photons detected did not allow for a stringent conclusion on the
265: possible variability of the source on longer time scales.
266:
267:
268: \section{Discussion\label{section:discussion}}
269:
270: We detected a {\it Chandra} source in the {\it ROSAT} error circle
271: (Augusteijn et al. 1997) and IPN ellipse (Hurley et al. 2000) of GRO
272: J1744--28, making it likely that we have detected the quiescent X-ray
273: counterpart of GRO J1744--28. The proposed infrared counterpart (Cole
274: et al. 1997; Augusteijn et al. 1997) is not consistent with our {\it
275: Chandra} position and the nature of this source is unclear (it might
276: be, as suggested by Augusteijn et al. 1997, an artifact, although Cole
277: et al. 1997 could not confirm this). Our quiescent flux of $2.5 - 5.3
278: \times 10^{-13}$ \funit~(0.5--10 keV; unabsorbed) is consistent with
279: the non-detection of GRO J1744--28 during the archival {\it ROSAT}
280: observation reported by Augusteijn et al. (1997; using their count
281: rate upper limit for GRO J1744--28, an 0.5--10 keV unabsorbed flux
282: upper limit range can be derived using PIMMS\footnote{Available at
283: http://heasarc.gsfc.nasa.gov/Tools/w3pimms.html} of
284: $7-10\times10^{-13}$ \funit, depending on the spectral model).
285:
286: One possible mechanism producing the quiescent emission is residual
287: accretion on to the surface of the neutron star. However, Cui (1997)
288: argued (based on the sudden decrease of the pulsation amplitude at
289: times when the 2--60 keV flux of GRO J1744--28, as measured with {\it
290: RXTE}, was below $\sim2\times 10^{-9}$ \funit) that GRO J1744--28 is
291: in the ``propeller'' regime at relatively low fluxes. In this regime,
292: the magnetic field of the neutron star inhibits accretion on to the
293: neutron star surface. Our {\it Chandra} flux for GRO J1744--28 is much
294: lower than the critical flux limit given by Cui (1997) strongly
295: suggesting that the source was in the propeller regime during our
296: observation. Therefore, accretion on to the surface of the neutron
297: star is not likely to be the cause behind the quiescent X-ray emission
298: of GRO J1744--28. Similar conclusions were also reached for the
299: quiescent emission from the transient X-ray pulsars A 0535+26
300: (Negueruela et al. 2000) and 4U 0155+63 (Campana et al. 2001).
301:
302: A possible mechanism to produce the observed quiescent X-rays when the
303: source is in the propeller regime might be accretion down to the
304: magnetospheric radius $r_{\rm m}$ (e.g., Stella et al. 1994; Corbet
305: 1996; Campana et al. 1998b), which is approximately give by $(GM_{\rm
306: ns})^{-1/7} \mu^{4/7} \dot{M}^{-2/7}$, in which $M_{\rm ns}$ is the
307: neutron star mass, $\mu$ the neutron star magnetic moment, and
308: $\dot{M}$ the accretion rate. The luminosity produced will be $L = G
309: M_{\rm ns} \dot{M}/r_{\rm m}$. Using our measured luminosity and the
310: magnetic field strength found by Cui (1997; $\sim2\times10^{11}$
311: Gauss), then $\dot{M} \sim 3 - 5 \times 10^{15}$ g s$^{-1}$ and
312: $r_{\rm m} \sim 2 - 3 \times 10^{8}$ cm. This radius is larger than
313: the corotation radius $r_{\rm c} = (GM_{\rm ns})^{1/3}(P_{\rm
314: spin}/2\pi)^{2/3}$ (with $P_{\rm spin}$ the neutron star spin period),
315: which is $\sim10^{8}$ cm for GRO J1744--28, and therefore it fulfills
316: the condition that the source should be in the propeller
317: regime. Although a rough estimate\footnote{Large uncertainties are
318: present in the exact luminosity of the source (due to uncertainties in
319: the spectral shape and in the distance to the source), its magnetic
320: field strength, and in certain unknown constants (like the accretion
321: efficiency), which have been assumed to be roughly 1. A full detailed
322: discussion of those uncertainties is beyond the scope of this {\it
323: Letter}.}, it indicates that accretion down to the magnetospheric
324: radius might be able to produce the quiescent X-ray luminosity of GRO
325: J1744--28. However, contributions to the X-ray luminosity might also
326: come from several other mechanisms, like accretion onto the neutron
327: star surface (possibly on localized areas such as the magnetic poles)
328: due to the leakage of matter through the magnetospheric barrier.
329:
330: Another contribution to the quiescent X-ray emission will be the
331: thermal X-ray emission from the neutron star surface which should give
332: a rock bottom lower limit on the X-ray luminosity. Due to the low
333: statistics of our data and the high column density towards GRO
334: J1744--28, we are not able to accurately probe such a thermal
335: component. However, if the temperature of the thermal component is
336: similar to what has been observed for the quiescent spectra of the
337: non-pulsating neutron star transients, then at least the detected
338: emission in our {\it Chandra} observation is mostly due to an another
339: component, which has its origin probably in one of the above discussed
340: mechanisms.
341:
342:
343: Progress in our understanding of the quiescent X-ray emission can be
344: made for GRO J1744--28 by obtaining longer {\it Chandra} or {\it
345: XMM-Newton} observations of this source. Such observations will be
346: able to constrain its luminosity and spectral shape much better than
347: we are able to due with the present {\it Chandra} data and will
348: determine if the quiescent spectrum of GRO J1744--28 is similar to
349: that of the quiescent non-pulsating systems, or that a fundamental
350: difference is present. Detecting the pulsations in observations with
351: sufficient time resolution will also constrain the mechanisms for the
352: quiescent X-rays in this system. Detecting the pulsations in
353: quiescence would also unambiguously proof that the detected source is
354: GRO J1744--28.
355:
356:
357: {\it Note added in manuscript:} After submission of our paper, we
358: became aware of the paper by Daigne et al. (2002) who reported on a
359: {\it XMM-Newton} observation of GRO J1744--28 in quiescence. Their
360: reported position and X-ray flux for the source are consistent with
361: ours.
362:
363: \acknowledgments
364:
365: RW was supported by NASA through Chandra Postdoctoral Fellowship grant
366: number PF9-10010 awarded by CXC, which is operated by SAO for NASA
367: under contract NAS8-39073. WQD is supported by the CXC grant
368: SAO-GO1-2150A. We thank Jon Miller for comments on a previous version
369: of this letter.
370:
371: \begin{references}
372:
373: \reference{}Aldcroft, T. L., Karovska, M., Cresitello-Ditmar, M. L.,
374: Cameron, R. A., \& Markevitch, M. L., 2000, Proc. SPIE, 4012, 650
375:
376: \reference{}Arnaud, K. 1996, in G. Jacoby \& J. Barnes (eds.), {\it
377: Astronomical Data Analysis Software and Systems V.}, Vol. 101, p. 17,
378: ASP Conf. Series.
379:
380:
381: \reference{}Asai, K., Dotani, T., Kunieda, H., Kawai, N. 1996, \pasj,
382: 48, L27
383:
384: \reference{}Asai, K., Dotani, T., Hoshi, R., Tanaka, Y., Robinson,
385: C. R., Terada, K. 1998, \pasj, 50, 611
386:
387: \reference{}Augusteijn, T. et al. 1997, \apj, 486, 1013
388:
389: \reference{}Bildsten, L. \& Rutledge, R. E. 2002, In: ``The Neutron
390: Star - Black Hole Connection'', (NATO ASI Elounda 1999),
391: eds. C. Kouveliotou, J. Ventura, \& E. van den Heuvel, p. 245
392:
393: \reference{}Brown, E. F., Bildsten, L., \& Rutledge, R. E. 1998, \apj,
394: 504, L95
395:
396: \reference{}Campana, S. \& Stella, L. 2000, \apj, 541, 849
397:
398: \reference{}Campana, S., Stella, L., Mereghetti, S., Colpi, M.,
399: Tavani, M., Ricci, D., Dal Fiume, D., Belloni, T. 1998a \apj, 499, L65
400:
401: \reference{}Campana, S., Colpi, M., Mereghetti, S., Stella, L.,
402: Tavani, M. 1998b, \aapr, 8, 279
403:
404: \reference{}Campana, S., Gastaldello, F., Stella, L., Israel, G. L.,
405: Colpi, M., Pizzolato, F., Orlandini, M., Dal Fiume, D. 2001, \apj,
406: 561, 924
407:
408: \reference{}Cash, W. 1979, ApJ, 228, 939
409:
410: \reference{}Cole, D. M. et al. 1997, 480, 377
411:
412: \reference{}Corbet, R. H. D., 1996, \apj, 457, L31
413:
414: \reference{}Cui, W. 1997, \apj, 482, L163
415:
416: \reference{}Daigne, F., Goldoni, P., Ferrando. P., Goldwurm, A.,
417: Decourchelle, A., Warwick, R. S. 2002, \aap, in press
418: (astro-ph/0202209)
419:
420: \reference{}Finger, M. H., Koh, D. T., Nelson, R. W., Prince, T. A.,
421: Vaughan, B. A., Wilson, R. B. 1996, \nat, 381, 291
422:
423: \reference{}Fishman, G., Kouveliotou, C., van Paradijs, J., Harmon,
424: B., Paciesas, W., Briggs, M., Kommers, J., \& Lewin, W. 1995, \iaucirc
425: 6272
426:
427:
428: \reference{}Hands, A., Warwick, R., Watson, M., \& Helfand, D. 2002,
429: In New Visions of the X-ray Universe in the XMM-Newton and Chandra
430: Era, 26--30 November 2001, ESTEC, The Netherlands, (astro-ph/0202180)
431:
432: \reference{}Hurley, K. et al. 2000, \apj, 537, 953
433:
434: \reference{}Kouveliotou, C., van Paradijs, J., Fishman, G. J., Brigss,
435: M. S., Kommers, J., Harmon, B. A., Meegan, C. A., Lewin,
436: W. H. G. 1996, \nat, 379, 799
437:
438: \reference{}Lewin, W. H. G., Rutledge, R. E., Kommers, J. M., van
439: Paradijs, J., Kouveliotou, C. 1996, \apj, 462, L39
440:
441: \reference{}Menou, K., Esin, A. A., Narayan, R., Garcia, M. R.,
442: Lasota, J.-P., McClintock, J. E. 1999, \apj, 520, 276
443:
444:
445: \reference{}Negueruela, I., Reig, P., Finger, M. H., Roche, P. 2000,
446: \aap, 356, 1003
447:
448: \reference{}Nishiuchi, M. et al. 1999, \apj, 517, 436
449:
450: \reference{}Rutledge, R. E., Bildsten, L., Brown, E. F., Pavlov,
451: G. G., Zavlin, V. E., 2001, \apj, 551, 921
452:
453: \reference{}Stella, L., Campana, S., Colpi, M., Mereghetti, S.,
454: Tavani, M. 1994, \apj, 423, L47
455:
456: \reference{}Wang, Q. D., Gotthelf, E. V., \& Lang, C. C. 2002, \nat,
457: 415, 148
458:
459: \reference{}Woods, P. M. et al. 1999, \apj, 517, 431
460:
461: \reference{}Zavlin, V. E., Pavlov, G. G., \& Shibanov, Yu. A., 1996,
462: \aap, 315, 141
463:
464: \end{references}
465:
466:
467: \begin{figure}
468: \begin{center}
469: \begin{tabular}{c}
470: \psfig{figure=f1.eps,width=16cm}
471: \end{tabular}
472: \figcaption{The {\it Chandra}/ACIS-I image (for photons with energy of
473: 1 keV or higher) of GRO J1744--28, rebinned by a factor of 2. Due to
474: the point-spread-function for off-axis point sources, the source
475: appears to be extended, but it is consistent with being a point
476: source. The coordinates are for epoch J2000.0. Also shown are the
477: {\it ROSAT} error circle (Augusteijn et al. 1997), the IPN
478: localization ellipse (Hurley et al. 2000) and the error circles of the
479: possible infrared counterpart (Augusteijn et al. 1997: large; Cole et
480: al. 1997: small). It is obvious that the {\it Chandra} source is
481: consistent with the {\it ROSAT} and the IPN source, but not with the
482: infrared source.\label{fig:image} }
483: \end{center}
484: \end{figure}
485:
486: \begin{figure}
487: \begin{center}
488: \begin{tabular}{c}
489: \psfig{figure=f2.eps,width=16cm,angle=-90}
490: \end{tabular}
491: \figcaption{The energy spectrum (for photons above 1 keV) of GRO
492: J1744--28 as measured with the {\it Chandra}/ACIS-I instrument. The
493: solid line through the data is the best blackbody fit.
494: \label{fig:spectra} }
495: \end{center}
496: \end{figure}
497:
498:
499:
500:
501:
502: \end{document}
503:
504:
505: