1: \documentclass[12pt,preprint]{aastex}
2:
3:
4:
5: \usepackage[]{natbib}
6:
7: \newcommand{\rxte}{{\it RXTE}}
8:
9:
10: \newcommand{\tfe}{1E~1048.1--5937}
11: \newcommand{\tfn}{1E~2259.1+586}
12: \newcommand{\soelong}{RXS~J170849.0--400910}
13: \newcommand{\soe}{1RXS~1708--4009}
14: \newcommand{\oft}{4U~0142+61}
15: \newcommand{\efo}{1E~1841--045}
16: \newcommand{\axj}{AX~J1845.0--0258}
17: \newcommand{\ett}{XTE~J1810--197}
18: \newcommand{\cxo}{CXOU~J0110043.1--721134}
19:
20:
21: \def\lapp{\ifmmode\stackrel{<}{_{\sim}}\else$\stackrel{<}{_{\sim}}$\fi}
22: \def\gapp{\ifmmode\stackrel{>}{_{\sim}}\else$\stackrel{>}{_{\sim}}$\fi}
23:
24:
25: \begin{document}
26:
27: \title{Anomalous X-ray Pulsar 1E~1048.1--5937: Pulsed Flux Flares and Large Torque Variations}
28:
29:
30:
31: \author{Fotis P. Gavriil\altaffilmark{1} and Victoria M. Kaspi\altaffilmark{1,2} }
32:
33: \altaffiltext{1}{Department of Physics,
34: McGill University, Rutherford Physics Building, 3600 University Street, Montreal, QC, H3A 2T8, Canada.}
35: \altaffiltext{2}{Department of Physics and Center for Space Research,
36: Massachusetts Institute of Technology, Cambridge, MA 02139}
37:
38:
39:
40: \begin{abstract}
41: We report on continued monitoring of the Anomalous X-ray pulsar (AXP)
42: \tfe\ using the \textit{Rossi X-ray Timing Explorer}. We confirm that
43: this pulsar has exhibited significant pulsed flux variability. The
44: principal features of this variability are two pulsed X-ray
45: flares. Both flares lasted several months and had well-resolved
46: few-week--long rises.
47: The long rise times of the flares are a
48: phenomenon not previously reported for this class of object.
49: The epochs of the flare peaks were MJD $52,218.8 \pm 4.5$ and $52,444.4
50: \pm 7.0$. Both flares had shorter rise than fall times. The flares
51: had peak fluxes of $2.21 \pm 0.16$ and $3.00 \pm 0.13$ times
52: the quiescent value. We estimate a total
53: 2--10~keV energy release of $\sim 2.7\times 10^{40}$~ergs and $\sim
54: 2.8\times 10^{41}$ ergs for the flares, assuming a distance of 5 kpc.
55: We also report large (factor of $\sim$12) changes
56: to the pulsar's spin-down rate on time scales of weeks to months, shorter
57: than has been reported previously. We find marginal evidence for
58: correlation between the flux and spin-down rate variability, with
59: probability of nonrandom correlation 6\%. We discuss the implications
60: of our findings for AXP models.
61: \end{abstract}
62:
63:
64: \keywords{pulsars: general ---pulsars: individual (\tfe)---X-rays: general }
65:
66:
67:
68: \section{Introduction}
69: \label{sec:intro}
70:
71: Anomalous X-ray pulsars (AXPs) are an exotic manifestation of young
72: neutron stars. AXPs are known for their steady, soft X-ray pulsations in
73: the period range of 6--12~s. The detection of X-ray bursts from
74: two AXPs has confirmed the common nature of these objects with that of
75: soft gamma repeaters \citep{SGRs; gkw02,kgw+03}, another exotic type of young
76: neutron star. Both classes of objects are believed to be magnetars,
77: i.e., powered by the decay of an ultrahigh magnetic field that has
78: a magnitude of $10^{14}$--$10^{15}$ G on the stellar surface. For recent
79: AXP reviews, see \citet{kg04a} and \citet{kas04}.
80:
81: One issue in AXP research has been flux stability. Historically,
82: two AXPs have been reported to be highly flux variable. \citet{opmi98}
83: collected all published flux measurements for AXP \tfe\ and concluded
84: that its total flux varies by as much as a factor of 10 between observations
85: spaced by typically 1--2 yr over $\sim$20 yr. Those data were from a
86: diverse set of instruments, including imaging and nonimaging
87: telescopes. Similarly, flux variability by a factor of greater than 4 was
88: reported for AXP \tfn\ by \citet{bs96}, using data also from a variety
89: of instruments.
90:
91: However, long-term {\it Rossi X-Ray Timing Explorer (RXTE)} monitoring
92: of the pulsed flux of \tfe\ by
93: \citet{kgc+01} and of \tfn\ by \citet{gk02} using a single instrument
94: and set of analysis software showed no evidence to support such large
95: variability.\footnote{Total flux measurements with \rxte\ were difficult
96: given the large field of view of the PCA and the low
97: count rates for the AXPs relative to the background.}
98: Also, \citet{tgsm02}, following a short {\it
99: XMM-Newton} observation of \tfe, compared the observed flux with those
100: measured by two other imaging instruments, {\it ASCA} and {\it
101: BeppoSAX}. They found that in the three observations, the total flux
102: was steady to within $\sim$30\%--50\%. They argued that the nonimaging
103: detections included in the \citet{opmi98} analysis may have been
104: contaminated by other sources in the instruments' fields of view; in
105: particular, the bright and variable X-ray source $\eta$ Carina lies only
106: 38$'$ away.
107:
108: A possible solution to this puzzle came with the discovery of a large
109: (greater than 10 times) long-lived flux enhancement from \tfn\ at the time of a
110: major outburst in 2002 June 18. This event was accompanied by many other
111: radiative changes as well as by a large rotational spin-up
112: \citep{kgw+03,wkt+04}. This suggests that past flux variability
113: reported in AXPs could be attributed to similar outbursts that went
114: undetected.
115:
116: We report here, using data from our continuing {\it RXTE} monitoring
117: program, the discovery of significant pulsed flux variability in
118: \tfe. This variability is mainly characterized by two long-lived pulsed
119: flux flares, having well-resolved rises a few weeks long. These are unlike
120: any previously seen flux enhancements in AXPs and SGRs and thus likely
121: represent a distinct physical phenomenon. We find no evidence for any
122: major associated bursting behavior. We also report large variations
123: in the spin-down torque on timescales of a few weeks/months. We find only
124: a marginal correlation between the flux and torque variations. We
125: argue that this poses another significant challenge to any
126: disk-accretion model for AXPs, but is not inconsistent with the
127: magnetar model.
128:
129: \section{Analysis and Results}
130: \label{sec:obs}
131:
132: All observations reported here were obtained with the Proportional
133: Counter Array \citep[PCA][]{jsss+96} on board \rxte. The timing
134: observations described below are a continuation of those reported by
135: \citet{kgc+01}. We refer the reader to that paper for details of the
136: analysis procedure. This \rxte\ monitoring program has shown that,
137: in general, AXPs have sufficient stability for phase coherent timing
138: \citep[see][for a review]{kg04a}. \tfe\ is an exception. For this
139: pulsar, we have achieved phase-coherent timing only over relatively
140: short data spans. In 2002 March, we adopted the strategy of
141: observing this source every week with three short ($\sim$2~ks)
142: observations. These closely spaced observations allow us to measure
143: the spin frequency with high precision weekly without phase connecting
144: over long baselines. This therefore allows us to determine the
145: spin-down rate with interesting precision on timescales of a few weeks.
146: Figure~\ref{fig:spin flux spectra}A shows the long-term spin history of
147: \tfe\ as measured by {\it RXTE}.
148:
149:
150:
151: Figure~\ref{fig:fit}A shows the spin-down rate $\dot{\nu}$
152: as a function of time over the interval for which we can make this
153: measurement. Plotted values of $\dot{\nu}$ were calculated by
154: measuring the slopes of each 5 adjacent values of
155: ${\nu}$. Note how $\dot{\nu}$ clearly varies greatly during our
156: observations, on all timescales to which we are sensitive.
157: From MJD 52,400 to MJD 52,620 $\dot{\nu}$ had changed by a factor of $\sim$12.
158: During the
159: $\sim$120 day interval from MJD 52,620 through MJD 52,740, $\dot{\nu}$ was a
160: factor of $\sim$4 larger than the long-term average spin-down
161: ($\langle \dot{\nu} \rangle=-6.48\times 10^{-13}\ \mathrm{Hz}\ \mathrm{s}^{-1}$). This
162: was followed by an abrupt decrease in magnitude by a factor of $\sim$2,
163: which was not resolved, and by subsequent additional variations.
164: At no time did we observe any episode of spin-up.
165:
166: We also monitor the pulsed flux of this source. In this analysis, data
167: from each observing epoch were also folded at the optimal pulse
168: period.
169: We calculated the rms pulsed flux using the method described
170: by \citet{wkt+04}.\footnote{In eq. (1) of \citet{wkt+04} there is a typographical error of a factor of 2 missing from the coefficients $\alpha_k$ and $\beta_k$; similarly a factor of 4 is missing from their respective variances. The calculations in that paper however, used the correct form of the equation.}
171: Given \tfe's highly sinusoidal pulse profile we used only
172: the first two harmonics to calculate the pulsed flux.
173: This method of
174: measuring flux is different from the one used in \citet{kgc+01} which
175: involved fitting a spectral model to extract a pulsed flux in cgs
176: units. Given the short length of the observations, fitting a spectral
177: model to the individual observations was not practical.
178: The pointing for two observations was slightly off-source
179: so we had to correct for reduced collimator response.
180: The pointing was on-source for all other observations.
181: Figure~\ref{fig:spin flux spectra}B shows our pulsed flux time series
182: in the 2--10~keV band. Pulsed flux time series in the 2--4 and
183: 4--6~keV bands look similar.
184:
185:
186: The pulsed flux time series clearly has significant structure. The
187: most obvious features are two long-lived flares. The first flare was
188: smaller and shorter lived than the second. The latter clearly
189: displayed significant structure in its decay. In estimating the following flare
190: properties, we define the first flare as having occurred between MJDs
191: 52,198 and 52,318 and the second having started on MJD 52,386, and we
192: take its end to be our last observation on MJD 53,030, although it
193: clearly has not yet ended (see Fig.~\ref{fig:fit}). We estimate that
194: the first flare had a peak flux of $2.21\pm 0.16$ times the quiescent pulse flux,
195: with the peak occurring at MJD $52,218.8 \pm 4.5$. Its rise time was $20.8
196: \pm 4.5$ days, and its fall time $98.9 \pm 4.5$ days. The second flare
197: peak was on MJD 52,444.4$\pm$7.0, and had a peak value of $3.00 \pm 0.13$
198: times the quiescent pulsed flux. Its rise time was $58.3 \pm
199: 7.0$~days, and its fall time is greater than 586 days. We estimate 2--10~keV
200: fluences of $(111 \pm 12)\times
201: 10^4\ \mathrm{counts}\ \mathrm{PCU}^{-1}$ and $(1136 \pm 38)\times
202: 10^4\ \mathrm{counts}\ \mathrm{PCU}^{-1}$ for the first and second flare,
203: respectively. \citet{tgsm02} measured a total flux in the
204: 2--10~keV energy range of $\sim 5 \times 10^{-12}$ ergs cm$^{-2}$~s$^{-1}$ and a
205: pulsed fraction of $\sim94\%$ (for energies $>$2 keV) from
206: \textit{XMM-Newton} observations of \tfe. This information, along with
207: our measured quiescent pulsed flux, allows us to scale our fluences to
208: estimate the total energy released in each flare. Assuming a distance
209: of 5 kpc \citep[see discussion in ][]{opk01}, we find a total energy
210: release of $\sim 2.7 \times 10^{40}$ ergs for the first flare and $\sim
211: 2.8\times 10^{41}$ ergs for the second flare, both in the 2--10 keV
212: band.
213:
214: Although we clearly detect both large flux variations and large changes
215: in the spin-down rate, the correlation between the two is marginal.
216: The Spearman rank order correlation coefficient $r_S = 0.28$, where 0
217: indicates no correlation and 1 indicates total correlation. The
218: probability of obtaining this value of $r_S$ or higher by random chance
219: is 6\%. Thus, there is marginal evidence of some correlation,
220: equivalent to a $\gapp 2\sigma$ result. From Figure~2, it is clear why
221: any correlation is not strong: for example, $\dot{\nu}$ changes very
222: little during the rise of the second flare, in the interval MJD
223: 52,380-52,420. Also, there is no short-term flux change when
224: $\dot{\nu}$ suddenly reaches its maximum absolute value (near MJD
225: 52,620), nor when it abruptly changes by a factor of $\sim$2 around MJD
226: 52,740.
227:
228: Hardness ratios (HRs) were measured by comparing the pulsed flux, as
229: measured by the method described above, in the 2--4 keV band to that in
230: the 4--6 keV band. Figure~\ref{fig:spin flux spectra}C shows our HR
231: measurements. The mean HR is 0.78. There is evidence for
232: spectral variability. The reduced $\chi^2$ of the HR time series is
233: 3.6 for 143 degrees of freedom. However, there is no evidence for any
234: correlation of HR with pulsed flux or torque. Our uncertainties,
235: however, are quite large; monitoring observations with an imaging
236: instrument would improve this situation.
237:
238: Intriguingly, the peak of the first flare was coincident with the
239: epochs during which we observed two SGR-like X-ray bursts from the
240: direction of this source in 2001
241: \citep[][indicated by arrows in Fig.~\ref{fig:spin flux spectra}]{gkw02}.
242: However, we found no other
243: SGR-like
244: bursts in any of the remaining data. For a detailed description of our
245: burst-searching algorithm see \citet{gkw04}. We also searched our
246: folded time series for pulse morphology variations using the method
247: detailed by \citet{gk02}. We find no evidence for significant pulse
248: profile changes at any epoch in our data set.
249:
250: \section{Discussion}
251: \label{sec:discussion}
252:
253: The long-lived flux enhancements with well-resolved rises that we have
254: observed in \tfe\ are very different from previously detected X-ray
255: flux variations in AXPs and SGRs, which show very abrupt rises
256: associated with major outbursts \citep[e.g.][]{kgw+03,wkt+04}.
257: The long-lived flux decay in those
258: sources has been attributed to burst afterglow, which is a cooling of the
259: crust following an impulsive heat injection from magnetospheric bursts
260: \citep{let02}. The much more gradual flux rises that we have observed in
261: \tfe\ comprise a new phenomenon not yet observed in any other AXP,
262: despite several years of careful and frequent {\it RXTE} monitoring.
263: These flux variations may provide a new diagnostic of the physical
264: origin of the persistent nonthermal emission in SGRs and AXPs, since
265: they are not contaminated by burst afterglow.
266: Also interesting are the large variations in spin-down rate
267: or torque. Torque variations by nearly a factor of 5 were already
268: reported from {\it RXTE} observations \citep{kgc+01}, on timescales of years.
269: Here we have shown that the torque can change
270: by at least a factor of $\sim$2 more, and on much shorter timescales,
271: namely, a few weeks to months.
272:
273: In considering the observed pulsed flux and torque variations, whether
274: they are correlated is an important issue. Our weekly monitoring of
275: the source unfortunately commenced only after most of the first flare
276: decayed. Prior to that, the monthly observations, taken in the form of
277: brief snapshots, did not allow anything about the rotational behavior
278: of the source to be determined when phase-coherent timing was not
279: possible. This was the case during the first flare. During the second
280: flare, the spin frequency was, interestingly, {\it most} stable during
281: the rise and peak of the flare. Furthermore, the stable spin-down
282: rate was at a lower magnitude than the long-term average.
283: Subsequently, $\sim 60$ days after the flux began to decay, the rate of
284: spin-down began to increase. Given timing observations during only one
285: flare, it is unclear whether these features are coincidences or not. However,
286: there is no strong evidence to support otherwise; similar torque
287: variations were seen in the past and were not accompanied by any
288: flaring (see Fig.~\ref{fig:spin flux spectra}).
289: Significant torque variations unaccompanied by severe
290: flux variability have been noted for \tfe\ prior to our \rxte\ monitoring
291: \citep[e.g.][]{pkdn00}.
292: Nevertheless, statistically, the probability that
293: they are uncorrelated is only 4\%; studying Figure~\ref{fig:fit}
294: suggests that if anything, slope transitions are correlated, if not the
295: slopes between transitions. Continued \rxte\ monitoring will help
296: identify any true correlations, particularly if the source exhibits
297: more variability.
298:
299: Can the magnetar model explain such behavior? The persistent emission in
300: magnetars has a spectrum that is well described by a two-component
301: model, consisting of a blackbody plus a hard power-law tail.
302: The thermal component is thought to arise from heat resulting from the
303: active decay of a high internal magnetic field \citep{td96a}; however,
304: thermal X-ray flux changes are not expected on as short a time scale as
305: we have measured in the absence of major bursts.
306: \citet{tlk02} put forth a model in which the
307: nonthermal component arises from resonant Compton scattering of
308: thermal photons by currents in the magnetosphere. In magnetars, these
309: currents are maintained by magnetic stresses acting deep inside its
310: highly conducting interior, where it is assumed that the magnetic field
311: lines are highly twisted. These magnetospheric currents in turn twist
312: the external dipolar field in the lesser conducting magnetosphere.
313: These magnetic stresses can lead to sudden outbursts or more gradual
314: plastic deformations of the rigid crust, thereby twisting the
315: footpoints of the external magnetic field and inducing X-ray luminosity
316: changes. The persistent non-thermal emission of AXPs is explained in
317: this model as being generated by these currents through
318: magnetospheric Comptonization and surface back-heating
319: \citep{td96a,tlk02}. Changes in X-ray luminosity, spectral hardness,
320: and torque have a common physical origin in this model and some
321: correlations are expected. Larger twists correspond to harder
322: persistent X-ray spectra, as is observed, at least when comparing the
323: harder SGR spectra to those of the softer AXPs. As noted by
324: \citet{kgc+01}, \tfe's hard photon spectral index ($\Gamma = 2.9$)
325: suggests that it is a transition object between the AXPs ($\Gamma \simeq$3--4)
326: and the SGRs ($\Gamma = $2.2--2.4). Hence, if during the flares
327: \tfe's magnetosphere was twisted to the SGR regime, we
328: expect spectral index variations of $\sim 0.5$. Spectral measurements
329: of such precision are not feasible with our short \rxte\ monitoring
330: observations.
331:
332: Decoupling between the torque and the luminosity can be
333: accounted for in the magnetar model. According to \citet{tlk02} the
334: torque is most sensitive to the current flowing on a relatively narrow
335: bundle of field lines that are anchored close to the magnetic pole, and
336: so only a broad correlation in spin-down rate and X-ray luminosity
337: is predicted, and in fact is observed for the
338: combined population of SGRs and AXPs \citep{mw01,tlk02}. However, for a
339: single source, whether an X-ray luminosity change will be accompanied
340: by a torque change depends on where in relation to the magnetic pole
341: the source of the enhanced X-rays sits. Similarly, large torque variations,
342: as we have observed, may occur in the absence of luminosity changes
343: if the former are a result of changes in the currents flowing only
344: in the small polar cap region.
345:
346: Note that energetically, the total release in these flares is
347: comparable to, although somewhat less than, that in the afterglows seen in
348: SGRs and in AXP \tfn\ \citep[see][for a summary]{wkt+04}. It
349: easily can be accounted for given the inferred magnetic energy of the
350: star.
351:
352: Although the magnetar model for AXPs has been spectacularly successful
353: in explaining their most important phenomenology, the anomalous behavior noted
354: for \tfe\ raises the possibility that perhaps it has a physical nature
355: different from other AXPs. It has also been suggested
356: that AXPs might be powered by accretion from fossil disks \citep{chn00,
357: alp01}. An increase in luminosity $L_X$ can easily be explained in accretion
358: models by an increase in the mass accretion rate $\dot{M}$, given that
359: $L_X \propto \dot{M}$. Transient changes in $\dot{M}$ are perhaps not
360: unreasonable to expect in fossil disk models, given the huge variations
361: seen in $\dot{M}$ of conventional accreting sources. However, in an
362: accretion scenario, we expect correlations
363: between luminosity and torque. In conventional disk-fed accreting pulsars
364: undergoing spin-up, one expects $\dot{\nu} \propto L_X^{6/7}$.
365: Such a correlation is seen approximately in
366: accreting pulsars, with discrepancies possibly attributable to changed
367: beaming or improper measurement of bolometric luminosities, the former
368: due to pulse profile changes, and the latter due to finite
369: bandpasses \citep{bcc+97}. As discussed by \citet{kgc+01}, for a
370: source undergoing regular {\it spin-down} as in \tfe, the prediction is less
371: clear; the form of the correlation depends on the unknown functional
372: form of the torque. For the propeller torque prescription of
373: \citet{chn00}, we find that $L_X\propto \dot{\nu}^{7/3}$, a much stronger
374: correlation than in the conventional spin-up sources. For a change
375: in $L_X$ by a factor of $\sim$3 as we have seen in the rise of the
376: second flare, we would expect
377: a simultaneous change in $\dot{\nu}$ by greater than $50\%$,
378: clearly ruled out by
379: our data. Conversely, for the abrupt change of $\dot{\nu}$ by a factor of
380: $\sim$2 (near MJD 52740), we expect a change in $L_X$ by a factor
381: of $\sim$5, definitely not seen. This appears to
382: pose a significant challenge to fossil-disk accretion models
383: for \tfe.
384:
385: Two infrared observations taken on MJD 52,324 \citep{ics+02}
386: and MJD 52372 \citep{wc02}
387: have shown that the IR counterpart
388: of this source is variable. However, the pulsed
389: X-ray flux at both those epochs was consistent with the quiescent value.
390: Furthermore, even though the X-ray flux has not yet returned to its
391: quiescent value, recent observations show that the source's proposed IR
392: counterpart is consistent with the fainter of the two previous observations
393: \citep{dk04}. This
394: decoupling between the IR and the X-ray flux contrasts with what
395: was observed in AXP \tfn, whose IR flux increased then decayed in
396: concert with the X-ray flux at the time of its 2002 outburst
397: \citep[][Tam et al. 2004, in preparation]{kgw+03}. This is
398: puzzling and suggestive of more than one mechanism for producing
399: IR emission in AXPs.
400:
401: \acknowledgements
402: We thank M. Lyutikov, S. Ransom, M.~Roberts, C. Thompson and
403: P.~Woods for useful discussions. This work is supported by NSERC
404: Discovery Grant 228738-03, NSERC Steacie Supplement 268264-03, a Canada
405: Foundation for Innovation New Opportunities Grant, FQRNT Team and
406: Centre Grants, and NASA Long-Term Space Astrophysics Grant NAG5-8063.
407: V.M.K. is a Canada Research Chair and Steacie Fellow. This research
408: has made use of data obtained through the High Energy Astrophysics
409: Science Archive Research Center Online Service, provided by the
410: NASA/Goddard Space Flight Center.
411:
412:
413:
414: \begin{figure}
415: \plotone{f1.eps}
416: \figcaption{Spin, flux, and spectral history of \tfe. \textit{(a)} Observed spin
417: frequencies vs. time. The points represent individual frequency
418: measurements. The solid lines represent the phase-connected intervals
419: as reported by \citet{kgc+01}. The dashed line is the long-term average
420: spin-down. \textit{(b)} Pulsed flux time series in the 2--10~keV
421: band. Arrows indicate the times at which the bursts reported by
422: \citet{gkw02} occurred. \textit{(c)} HR as a function of time. The
423: HRs displayed were computed for the pulsed flux in the energy range
424: (4--6~keV)/(2--4~keV).
425: \label{fig:spin flux spectra}}
426: \end{figure}
427:
428:
429:
430:
431: \begin{figure}
432: \plotone{f2.eps}
433: \figcaption{\textit{(a)} $\dot{\nu}$ vs. time over the interval for which
434: our data allow the measurement. The horizontal dashed line denotes the
435: long-term average spin-down rate, $\langle \dot{\nu} \rangle$. The right-hand scale is
436: the fractional difference of $\dot{\nu}$ and the long-term average spin-down rate.
437: \textit{(b)} Zoom-in of the pulsed flux time series in the 2--10~keV band.
438: Vertical dotted lines denote the chosen start and end ranges for characterizing
439: the two principal flares.
440: \label{fig:fit}}
441: \end{figure}
442:
443:
444: \bibliographystyle{apj}
445: %\bibliography{journals1,modrefs,psrrefs,crossrefs,myrefs}
446:
447: \begin{thebibliography}{23}
448: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
449:
450: \bibitem[{Alpar (2001)}]{alp01}
451: Alpar, M. A. 2001, ApJ, 554, 1245
452:
453: \bibitem[{Baykal \& Swank(1996)}]{bs96}
454: Baykal, A. \& Swank, J. 1996, ApJ, 460, 470
455:
456: \bibitem[{{Bildsten} {et~al.}(1997){Bildsten}, {Chakrabarty}, {Chiu}, {Finger},
457: {Koh}, {Nelson}, {Prince}, {Rubin}, {Scott}, {Stollberg}, {Vaughan},
458: {Wilson}, \& {Wilson}}]{bcc+97}
459: {Bildsten}, L., {Chakrabarty}, D., {Chiu}, J., {Finger}, M.~H., {Koh}, D.~T.,
460: {Nelson}, R.~W., {Prince}, T.~A., {Rubin}, B.~C., {Scott}, D.~M.,
461: {Stollberg}, M., {Vaughan}, B.~A., {Wilson}, C.~A., \& {Wilson}, R.~B. 1997,
462: ApJS, 113, 367
463:
464: \bibitem[{{Chatterjee} {et~al.}(2000){Chatterjee}, {Hernquist}, \&
465: {Narayan}}]{chn00}
466: {Chatterjee}, P., {Hernquist}, L., \& {Narayan}, R. 2000, ApJ, 534, 373
467:
468: \bibitem[{{Durant} \& {van Kerkwijk}(2004)}]{dk04}
469: {Durant}, M. \& {van Kerkwijk}, M.~H. 2004, in
470: Young Neutron Stars and Their Environments,
471: IAU Symposium 218, eds. B. Gaensler \& F. Camilo,
472: (San Francisco: Astronomical Society of the Pacific),
473: in press (astro-ph/0309801)
474:
475: \bibitem[{{Gavriil} \& {Kaspi}(2002)}]{gk02}
476: {Gavriil}, F.~P. \& {Kaspi}, V.~M. 2002, ApJ, 567, 1067
477:
478: \bibitem[{Gavriil {et~al.}(2002)Gavriil, Kaspi, \& Woods}]{gkw02}
479: Gavriil, F.~P., Kaspi, V.~M., \& Woods, P.~M. 2002, Nature, 419, 142
480:
481: \bibitem[{Gavriil {et~al.}(2004)Gavriil, Kaspi, \& Woods}]{gkw04}
482: ---. 2004, ApJ, in press; astro-ph/0310852
483:
484: %\bibitem[{Ibrahim {et~al.}(2004)Ibrahim, Markwardt, Swank, Ransom, Roberts,
485: % Kaspi, Woods, Safi-Harb, Balman, Parke, Kouveliotou, Hurley, \&
486: % Cline}]{ims+04}
487: %Ibrahim, A.~I., Markwardt, C., Swank, J., Ransom, S., Roberts, M., Kaspi,
488: % V.~M., Woods, P., Safi-Harb, S., Balman, S., Parke, W., Kouveliotou, C.,
489: % Hurley, K., \& Cline, T. 2004, ApJ, in press; astro-ph/0310665
490:
491: \bibitem[{{Israel} {et~al.}(2002){Israel}, {Covino}, {Stella}, {Campana},
492: {Marconi}, {Mereghetti}, {Mignani}, {Negueruela}, {Oosterbroek}, {Parmar},
493: {Burderi}, \& {Angelini}}]{ics+02}
494: {Israel}, G.~L., {Covino}, S., {Stella}, L., {Campana}, S., {Marconi}, G.,
495: {Mereghetti}, S., {Mignani}, R., {Negueruela}, I., {Oosterbroek}, T.,
496: {Parmar}, A.~N., {Burderi}, L., \& {Angelini}, L. 2002, ApJ, 580, L143
497:
498: \bibitem[{Jahoda {et~al.}(1996)Jahoda, Swank, Stark, Strohmayer, Zhang, \&
499: Morgan}]{jsss+96}
500: Jahoda, K., Swank, J., Stark, M., Strohmayer, T., Zhang, W., \& Morgan, E.
501: 1996, Proc. SPIE, 2808, 59
502:
503: \bibitem[{Kaspi(2004)}]{kas04}
504: Kaspi, V.~M. 2004, in Young Neutron Stars and Their Environments, {IAU}
505: Symposium 218, eds. B.~Gaensler \& F.~Camilo (San Francisco: Astronomical
506: Society of the Pacific), in press (astro-ph/0402175)
507:
508: \bibitem[{Kaspi \& Gavriil(2004)}]{kg04a}
509: Kaspi, V.~M. \& Gavriil, F.~P. 2004, in The Restless High-Energy Universe, ed.
510: E.~van~den Heuvel, J.~in't Zand, \& R.~Wijers (Elsevier), in press;
511: astro-ph/0402176
512:
513: \bibitem[{Kaspi {et~al.}(2001)Kaspi, Gavriil, Chakrabarty, Lackey, \&
514: Muno}]{kgc+01}
515: Kaspi, V.~M., Gavriil, F.~P., Chakrabarty, D., Lackey, J.~R., \& Muno, M.~P.
516: 2001, ApJ, 558, 253
517:
518: \bibitem[{Kaspi {et~al.}(2003)Kaspi, Gavriil, Woods, Jensen, Roberts, \&
519: Chakrabarty}]{kgw+03}
520: Kaspi, V.~M., Gavriil, F.~P., Woods, P.~M., Jensen, J.~B., Roberts, M. S.~E.,
521: \& Chakrabarty, D. 2003, ApJ, 588, L93
522:
523: \bibitem[{{Lyubarsky} {et~al.}(2002){Lyubarsky}, {Eichler}, \&
524: {Thompson}}]{let02}
525: {Lyubarsky}, Y., {Eichler}, D., \& {Thompson}, C. 2002, ApJ, 580, L69
526:
527: \bibitem[{{Marsden} \& {White}(2001)}]{mw01}
528: {Marsden}, D. \& {White}, N.~E. 2001, ApJ, 551, L155
529:
530: \bibitem[{Oosterbroek {et~al.}(1998)Oosterbroek, Parmar, Mereghetti, \&
531: Israel}]{opmi98}
532: Oosterbroek, T., Parmar, A.~N., Mereghetti, S., \& Israel, G.~L. 1998, A\&A,
533: 334, 925
534:
535: \bibitem[{\"Ozel {et~al.}(2001){\"Ozel}, {Psaltis}, \& {Kaspi}}]{opk01}
536: \"Ozel, F., Psaltis, D. \& Kaspi, V. M. 2001, ApJ, 563, 255
537:
538: \bibitem[{Paul {et~al.}}(2000)]{pkdn00}
539: {Paul}, B., {Kawasaki}, M., {Dotani}, T., \& {Nagase}, F. 2000, ApJ, 537, 319
540:
541: \bibitem[{Thompson \& Duncan(1996)}]{td96a}
542: Thompson, C. \& Duncan, R.~C. 1996, ApJ, 473, 322
543:
544: \bibitem[{Thompson {et~al.}(2002)Thompson, Lyutikov, \& Kulkarni}]{tlk02}
545: Thompson, C., Lyutikov, M., \& Kulkarni, S.~R. 2002, ApJ, 574, 332
546:
547: \bibitem[{{Tiengo} {et~al.}(2002){Tiengo}, {G{\" o}hler}, {Staubert}, \&
548: {Mereghetti}}]{tgsm02}
549: {Tiengo}, A., {G{\" o}hler}, E., {Staubert}, R., \& {Mereghetti}, S. 2002,
550: A\&A, 383, 182
551:
552: %\bibitem[{Vasisht {et~al.}(2000)Vasisht, Gotthelf, Torii, \& Gaensler}]{vgtg00}
553: %Vasisht, G., Gotthelf, E.~V., Torii, K., \& Gaensler, B.~M. 2000, ApJ, 542, L49
554:
555: \bibitem[{{Wang} \& {Chakrabarty}(2002)}]{wc02}
556: {Wang}, Z. \& {Chakrabarty}, D. 2002, ApJ, 579, L33
557:
558: \bibitem[{Woods {et~al.}(2004)Woods, Kaspi, Thompson, Gavriil, Chakrabarty,
559: Marshall, Flanagan, Heyl, \& Hernquist}]{wkt+04}
560: Woods, P.~M., Kaspi, V.~M., Thompson, C., Gavriil, F.~P., Chakrabarty, D.,
561: Marshall, H.~L., Flanagan, K., Heyl, J., \& Hernquist, L. 2004, ApJ, 605, 378
562:
563: \end{thebibliography}
564:
565:
566:
567: \end{document}
568:
569:
570: