1: \documentclass{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: %\usepackage{psfig}
4: %\voffset 0.05in
5: \newcommand{\etal}{et al.}
6: \newcommand{\loe}{\stackrel{<}{\sim}}
7: \newcommand{\goe}{\stackrel{>}{\sim}}
8: \newcommand\xte{{\it RXTE\/}}
9: \newcommand\chandra{{\it Chandra}}
10: \newcommand\xmm{{\it XMM-Newton}}
11:
12: \def\psr{\hbox{XTE J1829$-$098}}
13: \def\lpv{\hbox{2MASS J182944.55$-$095120.3}}
14:
15: \slugcomment{To appear in The Astrophysical Journal}
16:
17: %\submitted{}
18: %\accepted{}
19:
20: \shorttitle{X-ray and IR Observations of XTE J1829$-$098}
21: \shortauthors{Halpern and Gotthelf}
22:
23: \begin{document}
24:
25: \title{X-ray Observations and Infrared Identification of the Transient\\
26: 7.8~s X-ray Binary Pulsar XTE J1829$-$098}
27:
28: \author{J. P. Halpern and E. V. Gotthelf}
29:
30: \affil{Columbia Astrophysics Laboratory,
31: Columbia University, 550 West 120th Street, New York, NY 10027}
32:
33: \begin{abstract}
34: \xmm\ and \chandra\ observations of the transient 7.8~s pulsar
35: \psr\ are used to characterize its pulse shape and spectrum, and
36: to facilitate a search for an optical or infrared counterpart.
37: In outburst, the
38: absorbed, hard X-ray spectrum with $\Gamma = 0.76\pm 0.13$
39: and $N_{\rm H} = (6.0 \pm 0.6) \times 10^{22}$~cm$^{-2}$
40: is typical of X-ray binary pulsars. The precise \chandra\
41: localization in a faint state leads to the identification of
42: a probable infrared
43: counterpart at R.A. = $18^{\rm h}29^{\rm m}43.\!^{\rm s}98$,
44: decl. = $-09^{\circ}51^{\prime}23^{\prime\prime}\!.0$ (J2000.0)
45: with magnitudes $K=12.7$, $H=13.9$, $I>21.9$, and $R>23.2$.
46: If this is a highly reddened O or B star, we estimate
47: a distance of 10~kpc, at which the maximum observed X-ray luminosity
48: is $2 \times 10^{36}$ ergs~s$^{-1}$, typical of Be X-ray transients
49: or wind-fed systems. The minimum observed luminosity is
50: $3 \times 10^{32}(d/10\ {\rm kpc})^2$ ergs~s$^{-1}$.
51: We cannot rule out the possibility
52: that the companion is a red giant.
53: The two known X-ray outbursts of \psr\
54: are separated by $\approx 1.3$ yr, which may be the orbital period
55: or a multiple of it, with the neutron star in an eccentric orbit.
56: We also studied a late M-giant long-period variable
57: that we found only $9^{\prime\prime}$ from the X-ray position. It
58: has a pulsation period of $\approx 1.5$~yr, but is
59: not the companion of the X-ray source.
60:
61: \end{abstract}
62: \keywords{pulsars: individual (\psr) --- stars: neutron ---
63: stars: AGB and post-AGB --- stars: individual (\lpv) ---
64: stars: variable: other --- X-rays: binaries}
65:
66: \section {Introduction}
67:
68: Scans of the Galactic plane in 2004 July-August with the
69: {\it Rossi X-ray Timing Explorer} \citep[\xte;][]{bra93}
70: Proportional Counter Array \citep[PCA;][]{jah97}
71: detected a new, transient 7.8~s X-ray pulsar (Markwardt et al. 2004).
72: Those authors noted that the source's hard X-ray
73: spectrum favors classification as a high-mass X-ray binary (HMXB)
74: rather than a transient Anomalous X-ray Pulsar (AXP), although a
75: low-mass X-ray binary (LMXB) pulsar similar to 4U~1626$-$67
76: could not be ruled out.
77: We determined that three pointings had been made previously at this
78: location by the {\it Newton X-ray Multi-Mirror Mission} (\xmm),
79: and that a bright source was present
80: in one of these unpublished observations from 2003 March.
81: A preliminary analysis
82: clearly revealed its identity with \psr\ by
83: virtue of its 7.8~s pulsations and X-ray spectral parameters
84: compatible with the \xte\ measured ones (Halpern \& Gotthelf 2004a).
85: The repeated outburst of the source, separated by non-detections in
86: \xmm\ observations at a level at least 3400 times fainter,
87: is typical behavior of a transient HMXB.
88:
89: In \S 2, we describe the \xmm\ observation of \psr\ in its
90: outburst state, as well as upper limits from non-detections.
91: In \S 3, we report \chandra\ observations that localize the
92: source in its faint state, followed by optical and infrared
93: imaging in \S 4 that identifies a reddened counterpart.
94: The still uncertain classification of \psr\ is discussed in \S 5.
95: In the Appendix, we present photometry and spectroscopy of
96: a bright, newly recognized long-period variable star that is only
97: $9^{\prime\prime}$ from the X-ray source.
98:
99: \begin{deluxetable*}{llcccc}[t]
100: \tablecaption{X-ray Observations of \psr}
101: \tablehead
102: {\colhead{Mission} & \colhead{Date} &
103: \colhead{$F_x(2-10\ {\rm keV})$\tablenotemark{a}} & \colhead{$\Gamma$} &
104: \colhead{$N_{\rm H}$} & \colhead{References} \\
105: & & \colhead{(ergs cm$^{-2}$ s$^{-1}$)} & &
106: \colhead{($10^{22}$ cm$^{-2}$)} &
107: }
108: \startdata
109: {\it XMM} & 2002 Mar 27 & $< 5 \times 10^{-14}$ &...&...& 1 \\
110: {\it XMM} & 2003 Mar 27 & $ 8.2 \times 10^{-11}$
111: & $0.76 \pm 0.13$\tablenotemark{b}
112: & $6.0 \pm 0.6$\tablenotemark{b} & 1 \\
113: {\it XMM} & 2003 Sep 13 & $< 5 \times 10^{-14}$ &...&...& 1 \\
114: \xte & 2004 May 22 -- Jun 10 & $< 4.8 \times 10^{-11}$ &...&...& 2 \\
115: \xte & 2004 Jul 30 & $1.7 \times 10^{-10}$ &...&...& 2 \\
116: \xte & 2004 Aug 8 & $1.0 \times 10^{-10}$ & 1.0 & 10 & 2 \\
117: \chandra & 2007 Feb 14 & $2.5 \times 10^{-14}$ &...&...& 1 \\
118: \chandra & 2007 May 24 & $2.8 \times 10^{-13}$ &...&...& 1
119: \enddata
120: \tablenotetext{a}{Absorbed flux, assuming 1~mCrab $= 2.4 \times 10^{-11}$
121: ergs~cm$^{-2}$~s$^{-1}$ to convert from \citet{mar04}.}
122: \tablenotetext{b}{Uncertainties are 68\% confidence
123: for two interesting parameters (see Fig.~\ref{xrayspectrum}).}
124: \tablerefs{(1) this work;
125: (2) Markwardt et al. (2004).}
126: \label{xraytable}
127: \end{deluxetable*}
128:
129: \section{XMM-Newton Observations}
130:
131: Three \xmm\ pointings that include the location of
132: \psr\ were obtained in 2002 and 2003 as part of the Galactic Plane Survey
133: \citep{han04}. We reprocessed them using the emchain and epchain
134: scripts under Science Analysis System (SAS) version
135: xmmsas\_20060628\_1801-7.0.0.
136: Table~\ref{xraytable} lists the basic results
137: of these observations.
138: An X-ray source corresponding to \psr\ was detected only during
139: the 2.6~ks observation of 2003 March 27 (see Fig.~\ref{xmmimage}),
140: while the 2000 March 27 and 2003 September 13 images, having
141: exposure times of 6.8~ks and 6.4~ks, respectively, are blank at this
142: position. Thus, variability by at least a factor of 3400 is
143: indicated. The position of the source is
144: R.A = $18^{\rm h}29^{\rm m}44.\!^{\rm s}10$,
145: decl. = $-09^{\circ}51^{\prime}24.\!^{\prime\prime}1$ (J2000.0)
146: with a nominal 90\% uncertainty radius of $3.\!^{\prime\prime}2$.
147: This falls within the \xte\ error circle, and we confirm the
148: identification with \psr\ by finding the corresponding periodic signal at
149: barycentric period $P = 7.840 \pm 0.004$~s,
150: consistent with $P = 7.82 \pm 0.05$~s from \xte\ \citep{mar04}.
151:
152: \begin{figure}[b]
153: \centerline{
154: \includegraphics[width=3.2in,angle=-90.0]{xte1829_xmm_img_contour.ps}
155: }
156: \caption{\small Portion of the \xmm\ image of 2003 March 2007
157: centered on the \xte\ error ellipse of \psr\ \citep{mar04}.
158: The EPIC-pn and two EPIC MOS images have been combined.
159: }
160: \label{xmmimage}
161: \end{figure}
162:
163: The 73.4~ms sampling of the European Photon Imaging Camera
164: \citep[EPIC-pn;][]{tur03} data amply resolves the pulsations.
165: Figure~\ref{xraypulse} shows the EPIC-pn pulse profiles at several
166: energies from 1.0--12~keV. The pulsed fraction is dependent on
167: energy, ranging from $\approx 35\%$ at 1~keV to $\approx 70\%$ at 10~keV.
168: We also fitted the spectra of 2003 March 27, jointly
169: from the EPIC-pn and the two EPIC MOS CCDs, to an absorbed
170: power law. The resulting fit is acceptable ($\chi_{\nu}^2 = 1.0$
171: for 68 degrees-of-freedom) as shown in
172: Figure~\ref{xrayspectrum}, yielding a hard power law of photon
173: index $\Gamma = 0.76 \pm 0.13$ and
174: $N_{\rm H} = (6.0 \pm 0.6) \times 10^{22}$~cm$^{-2}$
175: ($1\sigma$ uncertainties).
176: These results are typical of HMXBs. In this case,
177: the measured $N_{\rm H}$ exceeds the Galactic 21~cm \ion{H}{1}
178: column density of $1.8 \times 10^{22}$~cm$^{-2}$ in this direction
179: \citep{dic90},
180: suggesting that some of the absorption is intrinsic to the binary.
181: However, the molecular hydrogen equivalent of the CO column density
182: in this direction is $N_H = 2N_{H_2} \sim 5 \times 10^{22}$~cm$^{-2}$
183: \citep{cle86}
184: which, together with \ion{H}{1}, could account for the X-ray
185: measured column density, but only if \psr\ is considerably
186: farther than the Galactic center.
187:
188: Due to the high stellar density at the \xmm\ position of \psr\
189: and the possibility that the optical counterpart is highly
190: reddened, a definite optical/IR counterpart could not be
191: identified from the \xmm\ position alone (see Fig.~\ref{opticalimages}).
192: Two possible optical counterparts were suggested by
193: \citet{hal04b}, but these are $3.\!^{\prime\prime}6$
194: and $7.\!^{\prime\prime}7$
195: from the \xmm\ position, so neither was compelling.
196: Therefore, the astrometric accuracy of \chandra\ was brought to bear
197: on the problem.
198:
199: \section{Chandra Observations}
200:
201: Three 5~ks observations were scheduled
202: with the \chandra\ {\it Advanced Camera for Imaging and Spectroscopy}
203: \citep[ACIS-I;][]{bur97}, at intervals of 3 months,
204: in the expectation that at least one snapshot, or a combination of the three,
205: would detect enough photons from this transient source to measure its position
206: to sub-arcsecond accuracy and enable an optical/IR identification.
207: The first \chandra\ observation was performed on 2007 February 14,
208: and the second on 2007 May 24.
209: The first \chandra\ observation detected a
210: candidate source consisting of three photons
211: only $2^{\prime\prime}\!.1$ from the \xmm\
212: coordinates of \psr. We reversed the pixel randomization
213: that is applied by the standard data processing in order to
214: smooth out the gridded appearance of the image resulting from
215: the undersampling of the telescope PSF by the ACIS CCDs.
216: In this case, it is more useful to recover the most
217: precise positions of the three photons. They fall within
218: $0^{\prime\prime}\!.7$ of each other, and have an rms dispersion in radius
219: of $0^{\prime\prime}\!.6$, consistent with the \chandra\ PSF.
220: We conclude that these three photons represent a single,
221: real source (see below).
222:
223: \begin{figure}[b]
224: \centerline{
225: \includegraphics[width=3.2in]{f2.eps}
226: }
227: \caption{Folded pulse profiles of \psr\ from the \xmm\
228: EPIC-pn observation of 2003 March 27. Background is
229: subtracted and count rates are
230: normalized so that the average per bin is 1.
231: }
232: \label{xraypulse}
233: \end{figure}
234:
235: The second \chandra\ observation detected 35 photons
236: at a position identical with the first,
237: R.A. = $18^{\rm h}29^{\rm m}43.\!^{\rm s}97$,
238: decl. = $-09^{\circ}51^{\prime}23^{\prime\prime}\!.2$ (J2000.0),
239: leaving no doubt that this is the counterpart of \psr.
240: For each observation, the nominal 90\% position
241: uncertainty is $0^{\prime\prime}\!.6$, which is dominated by
242: the aspect solution. Although several more
243: faint sources are detected in the \chandra\ images, none have obvious
244: optical/IR counterparts that can be used to further
245: refine the absolute astrometry.
246:
247: Given the small number of photons detected in the first \chandra\
248: observation, we address here the probability that it is a real
249: detection, as opposed to a background fluctuation.
250: First, we note that the group of
251: three photons meets the expectation of a hard, absorbed source,
252: with energies of 2.3, 3.5, and 5.6~keV. In the $2-8$~keV band,
253: there are a total of 29 ``events'' in a $30^{\prime\prime}$
254: radius around the \chandra\ position of \psr, which we take as an estimate
255: of the background rate for the calculation of chance coincidence.
256: Three of these 29 events comprise the candidate source, and fall
257: within the PSF of the second \chandra\ detection.
258: The probability that the three source
259: photons are a chance coincidence is
260: then $\approx 29\times 28\times 27\times (0.6/30)^6 = 1.4 \times 10^{-6}$.
261: We conclude that there is no doubt that the three-photon source is a
262: detection, even as its flux is uncertain by $\sim 60\%$.
263:
264: \begin{figure}[t]
265: \centerline{
266: \includegraphics[width=2.2in,angle=-90.0]{f3.eps}
267: }
268: \caption{X-ray spectrum of \psr\ from the \xmm\
269: observation of 2003 March 27, fitted to
270: an absorbed power law. The upper curve is from the EPIC-pn,
271: and the lower curve is from the combined EPIC-MOS detectors.
272: Residuals are in units of $\sigma$.
273: {\it Inset\/}: Confidence contours of spectral parameters
274: are $1\sigma, 2\sigma, 3\sigma$ for two interesting parameters.
275: }
276: \label{xrayspectrum}
277: \end{figure}
278:
279: \begin{figure}[b]
280: \vspace{0.1in}
281: \centerline{
282: \includegraphics[width=1.65in]{f4a.eps}
283: \includegraphics[width=1.65in]{f4b.eps}
284: }
285: \caption{Combined images of the field of \psr\ from the MDM 1.3m
286: telescope.
287: {\it Left\/}: $R$-band, limiting magnitude 23.2.
288: {\it Right\/}: $I$-band, limiting magnitude 21.9.
289: North is up, and east is to the left.
290: Each image is $0.\!^{\prime}8 \times 0.\!^{\prime}8$.
291: The large circle is the \xmm\ 90\% localization, of radius
292: $3.\!^{\prime\prime}2$, and the small circle is the
293: \chandra\ 90\% localization, of radius $0.\!^{\prime\prime}6$.
294: The brightest star in the $I$-band image is \lpv,
295: a late M giant and LPV,
296: with optical spectrum shown in Figure~\ref{opticalspectrum}
297: and photometry presented in Table~\ref{lpvtable}
298: and Figure~\ref{lpvlightcurve}.
299: }
300: \label{opticalimages}
301: \end{figure}
302:
303: The hardness ratio of the combined \chandra\ source detections is
304: HR=$+0.9\pm0.1$. Here, HR=$(C_{\rm h}-C_{\rm s})/(C_{\rm h}+C_{\rm s})$,
305: where $C_{\rm s}$ and $C_{\rm h}$ are the counts in the
306: $0.5-2$ and $2-10$~keV bands, respectively. This is consistent
307: with the hard, absorbed \xmm\ spectrum, and indicates that the
308: neutron star is still accreting at this low level.
309: Assuming the \xmm\ spectral parameters, the \chandra\ count rates in the two
310: observations correspond to $2-10$~keV fluxes of
311: $2.5 \times 10^{-14}$ and
312: $2.8 \times 10^{-13}$ ergs~cm$^{-2}$~s$^{-1}$, respectively,
313: derived using the web-based simulator
314: PIMMS\footnote{http://heasarc.gsfc.nasa.gov/Tools/w3pimms.html}.
315: Fluxes this small are to be expected from \psr,
316: since in two out of three \xmm\ observations,
317: we found an upper limit of $< 5 \times 10^{-14}$ ergs~cm$^{-2}$~s$^{-1}$.
318: Thus, the flux seen from \psr\ has varied by a factor of $\approx 6800$.
319:
320: \begin{figure}[b]
321: \centerline{
322: \vspace{0.1in}
323: \includegraphics[width=3.2in]{H_low.ps}
324: }
325: \caption{$H$-band image of the field of \psr\ from the
326: MDM 2.4m telescope on 2005 June 17.
327: North is up, and east is to the left.
328: The image is $0.\!^{\prime}9 \times 0.\!^{\prime}9$.
329: The black circle is the \xmm\ 90\% localization, of radius
330: $3.\!^{\prime\prime}2$, and the white circle is the
331: \chandra\ 90\% localization, of radius $0.\!^{\prime\prime}6$.
332: The probable counterpart at the \chandra\ position
333: has $H=13.9$ and also $K=12.7$ (the latter image not shown).
334: The bright, saturated star is \lpv,
335: a late M giant and LPV,
336: with optical spectrum shown in Figure~\ref{opticalspectrum}
337: and photometry presented in Table~\ref{lpvtable}
338: and Figure~\ref{lpvlightcurve}.
339: }
340: \label{irimage}
341: \end{figure}
342:
343: \section{Optical and Infrared Observations}
344:
345: We observed the location of \psr\ on a dozen epochs from 2005 April
346: to 2007 March using the MDM Observatory's 1.3 m McGraw-Hill telescope
347: \citep{weh71}.
348: Images were taken in the $R$ and $I$ bands using a SITe $1024\times1024$,
349: backside illuminated CCD with $24\mu$m ($0.\!^{\prime\prime}5$) pixels.
350: \citet{lan92} standard stars were used to calibrate one set of images
351: obtained under photometric conditions, and this calibration was
352: transferred to all of the other images, many of which were not photometric.
353: Figure~\ref{opticalimages} shows combined sets of the better images.
354: An astrometric solution was derived in the reference
355: frame of the USNO-B1.0 catalog \citep{mon03} using 29 stars that
356: have an rms dispersion of $0.\!^{\prime\prime}3$ about the fit,
357: comparable to the stated uncertainties in the catalog.
358: The nearest star is $3.\!^{\prime\prime}6$ from the \xmm\
359: position, but inconsistent with the \chandra\ position.
360: Otherwise, the error circles are blank.
361: We derive limits of $R>23.2$ and $I>21.9$ at the \chandra\ position.
362:
363: Infrared images in the $H$ and $K$ bands
364: were obtained on the MDM 2.4 m Hiltner telescope on 2005
365: June 17 using the TIFKAM imager/spectrometer \citep{pog98}
366: having a new $1024\times1024$ HAWAII2 HgCdTe detector array
367: with $0.\!^{\prime\prime}3$ pixels.
368: An astrometric calibration was derived
369: using 125 stars from the Two Micron All Sky Survey (2MASS)
370: with an rms dispersion of
371: $0^{\prime\prime}\!.14$ about the fit, comparable
372: to their expected uncertainties.
373: Figure~\ref{irimage} shows the $H$-band image,
374: on which a star not visible in the optical falls within
375: $0^{\prime\prime}\!.2$
376: of the \chandra\ localization of \psr. The IR position is
377: R.A. = $18^{\rm h}29^{\rm m}43.\!^{\rm s}98$,
378: decl. = $-09^{\circ}51^{\prime}23^{\prime\prime}\!.0$ (J2000.0).
379: We note that this star
380: is visible in 2MASS images, but not measured
381: in the 2MASS catalog, evidently due to the glare of the
382: nearby bright star. (The special properties of the latter
383: are described in the Appendix).
384: We calibrated six stars in the field
385: using their 2MASS magnitudes, finding that they are internally
386: consistent to within $\pm 0.05$ magnitudes. The resulting
387: magnitudes of the candidate counterpart are $H=13.9$ and $K=12.7$.
388: Therefore, we consider this a likely companion that is highly
389: reddened by intervening dust ($R-K>10$).
390:
391: \begin{deluxetable}{lcc}
392: \tablecaption{Optical Photometry of \lpv }
393: \tablehead
394: {\colhead{Date} & \colhead{$R$} & \colhead{$I$}
395: }
396: \startdata
397: 2005 Apr 23 & $17.99 \pm 0.02$ & $15.01 \pm 0.02$ \\
398: 2005 Aug 27 & $16.85 \pm 0.02$ & $13.90 \pm 0.02$ \\
399: 2005 Aug 29 & $16.85 \pm 0.02$ & $13.89 \pm 0.02$ \\
400: 2006 Mar 31 & $19.48 \pm 0.07$ & $16.47 \pm 0.02$ \\
401: 2006 Apr 02 & $19.53 \pm 0.04$ & $16.52 \pm 0.02$ \\
402: 2006 Apr 23 & ... & $16.77 \pm 0.08$ \\
403: 2006 Apr 29 & ... & $16.80 \pm 0.08$ \\
404: 2006 May 25 & $19.64 \pm 0.02$ & $16.56 \pm 0.02$ \\
405: 2006 Jun 29 & $19.03 \pm 0.02$ & $16.10 \pm 0.02$ \\
406: 2006 Sep 17 & $18.15 \pm 0.02$ & $15.17 \pm 0.02$ \\
407: 2007 Feb 23 & $16.57 \pm 0.02$ & $13.63 \pm 0.02$ \\
408: 2007 Mar 3 & $16.69 \pm 0.02$ & ... \\
409: 2007 Mar 8 & $16.74 \pm 0.02$ & $13.76 \pm 0.02$ \\
410: \enddata
411: \label{lpvtable}
412: \end{deluxetable}
413:
414: \begin{figure}[t]
415: \centerline{
416: \hfil
417: \includegraphics[width=2.4in,angle=-90.0]{f6.eps}
418: \hfil
419: }
420: \caption{\small Optical spectrum of \lpv\
421: in the vicinity of \psr\
422: obtained on the MDM 2.4m telescope on 2005 May 6.
423: }
424: \label{opticalspectrum}
425: \end{figure}
426:
427: \begin{figure}[b]
428: \vspace{0.1in}
429: \centerline{
430: \hfil
431: \includegraphics[width=3.2in]{f7.eps}
432: \hfil
433: }
434: \caption{\small Photometry of \lpv\
435: in the vicinity of \psr\ obtained on the MDM 1.3m telescope.
436: Data are listed in Table~\ref{lpvtable}.
437: }
438: \label{lpvlightcurve}
439: \end{figure}
440:
441: \section{Discussion}
442:
443: Accepting the association with a bright infrared counterpart,
444: we rule out the hypothesis of an isolated
445: neutron star, specifically a transient AXP, whose periods fall
446: in the range 2--12~s, encompassing that of \psr.
447: The X-ray spectrum is also harder than that of any AXP, being similar
448: to high-mass binaries, and also to the LMXB pulsar
449: 4U~1626$-$67 \citep{kii86}, as noted by \cite{mar04}.
450:
451: It is unlikely that \psr\
452: is a Roche-lobe filling LMXB, as its ratio of IR to X-ray
453: luminosity is too large to be accounted for by either
454: accretion-disk reprocessing or X-ray heating of the companion star's
455: photosphere, indicating instead that the IR emission must be
456: intrinsic to the companion star. Quantitatively, if we
457: suppose that the IR/optical spectrum has the
458: form $F_{\nu} \propto \nu^{1/3}$,
459: typical of accretion disk emission, then the
460: intrinsic IR color is $(H-K)_0 = 0.44$,
461: requiring $A_K = 1.34$.
462: The intrinsic flux between the $B$ and $K$ bands is then
463: $\approx 1.4 \times 10^{-10}$ ergs~cm$^{-2}$~s$^{-1}$,
464: which is comparable to the X-ray flux in the high state.
465: Such optical/X-ray ratios are typical of Be X-ray binaries
466: or wind-fed HMXBs, as is the high-state
467: X-ray luminosity of $2 \times 10^{36}\ (d/10\ {\rm kpc})^2$
468: ergs~s$^{-1}$ (see below).
469:
470: However, we cannot rule out the possibility that
471: the companion star is a low-mass red giant,
472: similar to the pulsar GX~1+4, and other recently discovered
473: M-giant counterparts of X-ray sources \citep{mas07}.
474: In either the HMXB or the M-giant case, the infrared emission
475: comes from the intrinsic stellar photosphere, and is too bright
476: to be affected by the state of the X-ray source.
477:
478: The extreme amplitude of X-ray variability, by a factor
479: of 6800, is most typical of a Be X-ray transient,
480: or perhaps a wind-fed system in which the
481: neutron star is in an eccentric orbit. X-ray turn-ons
482: may occur when the neutron star passes through the
483: circumstellar disk around a Be star companion or the
484: denser parts of a stellar wind.
485: Although the sparse temporal sampling of this source does
486: not allow a meaningful estimate of its recurrence time,
487: its two known outbursts were separated by $\sim 1.3$~yr,
488: which may be the orbital period or a multiple of it.
489: %The orbital
490: %period is probably not shorter than $\sim 30$ days, since
491: %the high state detected by \xte\ during its scans and following
492: %pointed observation lasted at least 10 days,
493: %while a low state lasted at least 20 days \citep{mar04}.
494: The relation
495: between orbital and spin periods of Be/X-ray binaries
496: \citep{cor86,lay05} would suggest an orbital period
497: in the range 20--60 days.
498: More frequent X-ray monitoring, or pulse timing in a high
499: state, may be able to discover the orbital period.
500:
501: The IR counterpart of \psr\ appears too bright, given
502: its large extinction, to be a main-sequence B star. A star
503: of spectral type B0V has absolute magnitude $M_K=-2.01$ \citep{han97}
504: and $H-K=-0.08$ \citep{win97}. In this case, since we measure
505: $H=13.9$ and $K=12.7$, the required extinction $E(H-K)=1.28$
506: is equivalent to $A_V \approx 20$, which corresponds to
507: $d \sim 10$~kpc assuming $A_V/d \sim 2$ mag~kpc$^{-1}$.
508: At this distance, a B0V star would have apparent magnitude $K=15.2$.
509: However, Be stars are generally found 0.5--1 magnitudes above
510: the main sequence \citep{sle88}, and also up to luminosity
511: class III. So we do not rule out a Be star based on
512: this crude argument and approximate distance.
513: At $d=10$~kpc, the maximum observed 2--10~keV
514: X-ray luminosity of \psr\ is $2 \times 10^{36}$ ergs~s$^{-1}$,
515: typical of Be X-ray transients and wind-fed systems,
516: and the minimum luminosity
517: is $\sim 3 \times 10^{32}$ ergs~s$^{-1}$. At this distance, it
518: remains inconclusive whether absorption local to the X-ray source,
519: e.g., from the companion's wind, is needed to help account for the
520: X-ray measured column density.
521:
522: Although most transient HMXBs are of the Be type,
523: it is possible that the companion of \psr\
524: is an O or B supergiant. Many such supergiant
525: X-ray pulsars having large column densities are
526: being found by the {\it INTEGRAL} Observatory
527: \citep{wal06}. However, their spin periods are
528: generally longer than 100~s, while the 7.8~s period
529: of \psr\ is typical of a Be X-ray binary. One possible
530: counterexample to this period trend is the 4.7~s
531: pulsar AX~J1841.0$-$0536 \citep{bam01}, optically identified
532: by \citet{hal04} and classified as a B0.2Ib supergiant
533: by \citet{neg05,neg07} \citep[see also][]{nes07}.
534: Further observations are needed to classify \psr.
535:
536: \section {Conclusions}
537:
538: We have characterized the X-ray spectrum and amplitude of
539: variability of the poorly studied transient X-ray pulsar
540: \psr, showing that it emits hard X-rays, i.e., accretes,
541: over a range of at least 6800 in luminosity, and has
542: had at least two outbursts in a period of 1.3~yr.
543: From the \chandra\ position, a reddened ($R-K>10$)
544: IR counterpart was found that is too bright
545: to be either an AXP or a low-mass main sequence companion.
546: However, we cannot determine from our
547: $H$ and $K$ photometry alone the spectral type of
548: the companion, whether Be, OB supergiant, or red giant.
549: The large X-ray measured $N_{\rm H}$ in excess of the
550: total Galactic column suggests the presence of
551: absorption local to the source, e.g., the wind
552: or circumstellar disk of the companion star.
553: We estimated the distance to the source by assuming that
554: the companion is of spectral type B0, and translating
555: the $A_V=20$ implied by its $H-K$ color to $d \sim 10$~kpc.
556: This corresponds to a maximum X-ray luminosity of
557: $2 \times 10^{36}\ (d/10\ {\rm kpc})^2$ ergs~s$^{-1}$.
558: Most important, infrared spectroscopy is now needed to classify
559: the companion and obtain an actual measurement of its
560: distance.
561:
562: \acknowledgments
563:
564: We thank Ian McGreer for obtaining the crucial infrared data at
565: MDM Observatory, and Pietro Reviglio for obtaining the optical
566: spectrum shown in Figure~\ref{opticalspectrum}.
567: This investigation is based on observations obtained with \xmm, an ESA
568: science mission with instruments and contributions directly funded by
569: ESA Member States and NASA. Support for this work was provided by
570: the National Aeronautics and Space Administration through
571: {\it Chandra} Award SAO GO7-8035X
572: issued by the {\it Chandra} X-ray Observatory Center,
573: which is operated by the Smithsonian Astrophysical Observatory
574: for and on behalf of NASA under contract NAS8-03060.
575: This publication makes use of data products from the Two Micron All Sky
576: Survey, which is a joint project of the University of Massachusetts and
577: the Infrared Processing and Analysis Center/California Institute of
578: Technology, funded by the National Aeronautics and Space Administration
579: and the National Science Foundation. We thank the anonymous referee
580: for useful suggestions.
581:
582: %\clearpage
583:
584: \appendix
585:
586: \section{A Long-Period Variable in the Vicinity of XTE J1829$-$098}
587:
588: We originally suggested a very bright, red star,
589: \lpv, with 2MASS magnitudes $J=8.04$, $H=6.11$, and $K=4.98$
590: that is $7.\!^{\prime\prime}7$ from the \xmm\ position,
591: as a possible supergiant companion of \psr\ \citep{hal04b}.
592: It is the saturated star in the
593: infrared image of Figure~\ref{irimage}.
594: While we have now ruled it out, we did discover in the course
595: of this investigation that it is an
596: extreme late-type red giant or supergiant that displays
597: long-term photometric variability typical of such stars.
598:
599: Figure~\ref{opticalspectrum} shows a spectrum of \lpv\
600: obtained on the MDM 2.4m telescope on 2005 May 6.
601: No flux is detected from it shortward of 6800 \AA.
602: Longward of 6800~\AA\ are deep TiO and VO bands
603: indicating spectral type M9-10
604: \citep[see][for comparison spectra]{woo83b,flu94}.
605: Figure~\ref{lpvlightcurve} and Table~\ref{lpvtable}
606: show the photometric variability of this star in
607: our $R$ and $I$-band images, exceeding 3 magnitudes
608: with an apparent period of $\approx 1.5$~yr. In this
609: study, it was important that the same detector/filter
610: combination was used each time, since the complex
611: red spectrum of \lpv\ would yield
612: very different photometry through slightly different bandpasses.
613:
614: Red supergiant and asymptotic giant branch
615: (AGB) stars are long-period variables (LPVs: prototype Mira)
616: due to the pulsational instability of their envelopes, and they
617: can vary by several magnitudes on time scales of hundreds of days.
618: We conclude that this star is a luminous giant or supergiant
619: because its magnitudes indicate significant extinction,
620: even in the $K$ band. Its 2MASS color is $J-K=3$, while
621: LPVs generally have intrinsic $(J-K)_0=1.1-1.8$ \citep{woo83a}.
622: Adopting $(J-K)_0=1.5$, we estimate $E(B-V)\approx 3$,
623: $A_V \approx 9$, and thus $d\sim 4.5$~kpc assuming
624: $A_V/d \sim 2$ mag~kpc$^{-1}$.
625: In this case, the absolute $K$ magnitude is $M_K \sim -9.4$,
626: similar to AGB stars having
627: periods of $\approx 1.5$~yr \citep{woo83b}.
628:
629: Until a spectral type and distance to \psr\ are determined,
630: we have no strong reason to suppose that \lpv\ is a
631: member of a stellar association that includes \psr.
632:
633: \begin{thebibliography}{}
634:
635: \bibitem[Bamba \etal(2001)]{bam01}
636: Bamba, A., Yokogawa, J., Ueno, M., Koyama, K., \& Yamauchi, S.
637: 2001, \pasp, 53, 1179
638:
639: \bibitem[Bradt \etal(1993)]{bra93}
640: Bradt, H. V., Rothschild, R. E., \& Swank, J. H. 1993, \aaps, 97, 355
641:
642: \bibitem[Burke \etal(1997)]{bur97}
643: Burke, B. E., Gregory, J., Bautz, M. W., Prigozhin, G. Y., Kissel, S. E.,
644: Kosicki, B. N., Loomis, A. H., \& Young, D. J. 1997, IEEE Trans. Electron
645: Devices, 44, 1633
646:
647: \bibitem[Clemens \etal(1986)]{cle86}
648: Clemens, D. P., Sanders, D. B., Scoville, N. Z., \& Solomon, P. M.
649: 1986, \apjs, 60, 297
650:
651: \bibitem[Corbet(1986)]{cor86}
652: Corbet, R. H. D. 1986, \mnras, 220, 1047
653:
654: \bibitem[Dickey \& Lockman(1990)]{dic90}
655: Dickey, J. M., \& Lockman, F. J. 1990, \araa, 28, 215
656:
657: \bibitem[Fluks \etal(1994)]{flu94}
658: Fluks, M .A., Plez, B., The, P. S., de Winter, D., Westerlund, B. E., \&
659: Steenman, H. C. 1994, A\&AS, 105, 311
660:
661: \bibitem[Halpern \& Gotthelf(2004a)]{hal04a}
662: Halpern, J. P., \& Gotthelf, E. V. 2004a, ATel, 319, 1
663:
664: \bibitem[Halpern \& Gotthelf(2004b)]{hal04b}
665: --------. 2004b, ATel, 344, 1
666:
667: \bibitem[Halpern \etal(2004)]{hal04}
668: Halpern, J. P., Gotthelf, E. V., Helfand, D. J., Gezari, S., \&
669: Wegner, G. A. 2004, ATel, 289, 1
670:
671: \bibitem[Hands \etal(2004)]{han04}
672: Hands, A. D. P., Warwick, R. S., Watson, M. G., \& Helfand, D. J. 2004,
673: MNRAS, 351, 31
674:
675: \bibitem[Hanson \etal(1997)]{han97}
676: Hanson, M. M. Howarth, I. D., \& Conti, P. S. 1997, \apj, 489, 698
677:
678: \bibitem[Houdashelt \etal(2000)]{hou00}
679: Houdashelt, M. L., Bell, R. A., Sweigart, A. V., \& Wing, R. F. 2000,
680: AJ, 119 1424
681:
682: \bibitem[Jahoda \etal(1997)]{jah97}
683: Jahoda, K., Swank, J. H., Giles, A. B., Stark, M. J., Stohmayer, T.,
684: Zhang, W., \& Morgan, E. H. 1996, Proc. SPIE, 2808, 59
685:
686: \bibitem[Kii \etal(1986)]{kii86}
687: Kii, T., Hayakawa, S., Nagase, F., Ikegami, T., \& Kawai, N.
688: 1986, \pasj, 38, 751
689:
690: \bibitem[Landolt(1992)]{lan92}
691: Landolt, A. 1992, \aj, 104, 340
692:
693: \bibitem[Laycock \etal(2005)]{lay05}
694: Laycock, S., Corbet, R. H. D., Coe, M. J., Marshall, F. E., Markwardt, C.,
695: \& Lochner, J. 2005, \apjs, 161, 96
696:
697: \bibitem[Markwardt \etal(2004)]{mar04}
698: Markwardt, C. B., Swank, J. H., \& Smith, E. A. 2004, ATel, 317, 1
699:
700: \bibitem[Masetti \etal(2007)]{mas07}
701: Masetti, N., \etal\ 2007, \aap, 470, 331
702:
703: \bibitem[Monet \etal(2003)]{mon03}
704: Monet, D., \etal\ 2003, \aj, 125, 984
705:
706: \bibitem[Negueruela \etal(2005)]{neg05}
707: Negueruela, I., Smith, D. M., Reig, P., Chaty, S., \& Torrej\'on, J. M.
708: 2006, in Proc. The X-ray Universe 2005, ed. A. Wilson (ESA SP-604, Vol. 1;
709: Noordwijk: ESA), 165 (astro-ph/0511088)
710:
711: \bibitem[Negueruela \etal(2007)]{neg07}
712: Negueruela, I., Smith, D. M., Torrej\'on, J. M., \& Reig, P.
713: 2007, in Proc. Sixth INTEGRAL Workshop, The Obscured Universe (Moscow)
714: (arXiv:0704.3224)
715:
716: \bibitem[Nespoli \etal(2007)]{nes07}
717: Nespoli, E., Fabregat, J., \& Mennickent, R. 2007, ATel, 983, 1
718:
719: \bibitem[Pogge \etal(1998)]{pog98}
720: Pogge, R. W., \etal\ 1998, Proc. SPIE, 3354, 414
721:
722: \bibitem[Slettebak(1988)]{sle88}
723: Slettebak, A. 1988, \pasp, 100, 770
724:
725: \bibitem[Turner \etal(2003)]{tur03}
726: Turner, M. J. L., Briel, U. G., Fernando, P., Griffiths, R. G.,
727: \& Villa, G. E. 2003, Proc. SPIE, 4851, 169
728:
729: \bibitem[Walter \etal(2006)]{wal06}
730: Walter, R., \etal\ 2006, \aap, 453, 133
731:
732: \bibitem[Wehinger \& Mohler(1971)]{weh71}
733: Wehinger, P. A, \& Mohler, O. C. 1971, Sky \& Telescope, 41, 72
734:
735: \bibitem[Winkler(1997)]{win97}
736: Winkler, H. 1997, \mnras, 287, 481
737:
738: \bibitem[Wood \& Bessell(1983)]{woo83a}
739: Wood, P. R., \& Bessell, M. S. 1983, ApJ, 265, 748
740:
741: \bibitem[Wood \etal(1983)]{woo83b}
742: Wood, P. R., Bessell, M. S., \& Fox, M. W. 1983, ApJ, 272, 99
743:
744: \end{thebibliography}
745:
746: \end{document}
747: